Tag: media

Trending? Social Media Attention on Russia’s War in Ukraine

20240218 Social Media Attention on Russia Image 01

Russia’s invasion of Ukraine is one of the most important geopolitical events of the 21st century. For almost two years, international news outlets have been covering the war, often providing daily or even hourly updates. But what is the level of public interest and public engagement in countries around the world? When does the war capture an international audience’s attention and what are the events that supplant it? This brief uses data on X (formerly Twitter) trends in 62 countries to address these questions.

The competition for attention is a defining feature of our information landscape. From the relentless stream of social media updates to the myriad of news articles vying for our clicks, individuals are constantly bombarded with information, each competing for a slice of their limited attention. Amidst this cacophony of voices, certain topics rise to the forefront, capturing the collective consciousness and dominating public discourse.

Russia’s war in Ukraine has, for obvious reasons, commanded significant media coverage over the past two years. It has been described as a hybrid war, where conventional military tactics are increasingly combined with non-traditional methods. This includes an information war, fought with narratives to manipulate people’s perceptions, spread falsehoods, or enlist support. To a large extent, this information war has taken place on social media. On the one hand, social media platforms have been used to spread disinformation and propaganda. For example, we’ve seen the spread of false narratives about the causes of the war, the actions of the different parties involved, and the suffering of the Ukrainian people. But on the other hand, social media has also been used to counter this disinformation, with fact-checking initiatives and grassroots efforts to promote accurate information.

This policy brief analyses the prominence of the war in social media discourse. While the content on traditional media outlets provides a snapshot of the supply of information, platforms like X/Twitter offer a unique window into the broader population’s demand for that information and how they evolve over time. Whether or not hashtags related to Russia’s war in Ukraine are trending in a given country, depends not just on the public’s interest in the war relative to other events in the news, but also on the level of interest relative to sport, music, television, and cats. By tracking the prevalence of trending hashtags, we can gain insights into the public’s engagement with Russia’s war in Ukraine, going beyond traditional media narratives and high-level governmental discussions to uncover the conversations and sentiments that shape broader public opinion.

The X/Twitter data suggest that in most countries, social media attention in the Russian war on Ukraine has been short-lived and sporadic. On February 24, 2022, Ukraine-related hashtags were trending in 100 percent of the countries in our dataset. Two weeks later, on March 9, 2022, they were trending in only 3 percent of the countries. We find that geographical proximity to the conflict is a strong predictor of social media interest. Related hashtags trend most frequently in Eastern, Central and Northern Europe. We also document spikes in interest around events that link a country to the war in Ukraine: announcements of military assistance or visits by Ukraine’s President Volodymyr Zelenskyj. Finally, we compare the hashtags trending in NATO countries to those trending in countries that either sided with Russia or abstained from voting in a critical UN resolution and show significant differences between the two groups.

Data and Methodology

The source for our dataset is archive.twitter-trending.com – a website that records trending hashtags on X/Twitter across countries and over time. We scrape this website to collect (i) the five highest volume topics in each country on each day and (ii) the five longest-trending topics in each country on each day (these two categories can overlap). Our sample consists of the 62 countries available on the website and covers the timeframe July 2021 to December 2023. From this, we construct a country-by-day panel dataset with 55,862 observations.

We identify 11 topic categories that collectively account for the overwhelming majority of trending topics related to Russia’s war in Ukraine. These topics and their relative frequency are shown in Figure 1. The three dominant categories are “Ukraine”, “Russia” and ”Putin”. We use Google’s translation software to translate non-English tweets which account for a significant fraction of the dataset.

Figure 1. Frequency of hashtags in 11 category topics.

Note: This chart shows the number of times topics assigned to our 11 war-related categories were among the top five longest trending topics (in orange) or the top five highest volume topics (in blue) in any country on any day in our dataset. The source are data scraped from archive.twitter-trending.com

Figure 1 shows that it is more common for war-related topics to be among the highest volume topics on a given day than among the longest trending topics. This suggests that these topics attract a lot of interest in a narrow timeframe (e.g. when news breaks) but are relatively less likely to remain prominent over a whole day. Despite this difference, we find that the distinction between highest-volume and longest-trending does not affect any of the patterns we observe when comparing across countries or time. For simplicity, the results shown below all use the highest-volume measure.

It is important to acknowledge the limitations of the X/Twitter data. Firstly, the population actively using X/Twitter is not representative of the overall population. Secondly, the composition of users may differ across countries which complicates cross-country comparisons. Trending hashtags provide an indicator of public interest that is informative only because we do not have high frequency, nationally representative surveys that are comparable across countries. Finally, we are only able to observe the top-five hashtags in a country on any given day. In principle, a war-related topic could increase in absolute volume from one day to the next, while still being crowded out of the top five.

Geographic Variation in Attention

Social media attention to the war in Ukraine varies greatly across countries. The map in Figure 2 shows the proportion of days when any hashtag from the considered categories was among the top-five most tweeted, for each country in the database since the start of the war. Interest has, on average, been higher in Europe as well as in Anglo-Saxon countries. In contrast, other regions of the world exhibited less sustained interest, as indicated by the lower frequency of related hashtags among the top-five most tweeted topics.

Figure 2. Prevalence of war-related hashtags.

Note: The map shows the share of days on which war-related hashtags (in our 11 categories) were among the top five highest volume topics on X/Twitter between 24/02/2022 and 18/12/2023. Countries in white are not among the 62 countries in the dataset. The source are data scraped from archive.twitter-trending.com

To some extent, this heterogeneity is explained by distance. Figure 3 plots the frequency of war-related trends against geographical proximity to the conflict zone (represented by the distance from each country’s capital to the city of Kharkiv in eastern Ukraine, a major point of focus during the ongoing war). The relationship is clearly negative, suggesting that physical distance from the crisis reduces the intensity of online discourse and public interest. Unsurprisingly, the number of related trends is highest in countries directly or indirectly involved in the conflict – Ukraine, Russia, and Belarus – as well as in Latvia which borders both Russia and Belarus.

Figure 3. Frequency of war-related hashtags and distance from Kharkiv.

Note: The chart shows the log of the distance from each country’s capital city to the city of Kharkiv in km on the x-axis and the logged frequency of war-related topics among the top five highest volume topics in that country between 24/02/2022 and 18/12/2023 on the y-axis. The source are data scraped from archive.twitter-trending.com

Variation in Attention Over Time

Over the past two years, the war has sustained a relatively high intensity. By contrast, global attention on X/Twitter has been more sporadic, spiking around specific events. This is shown in Figure 4, which plots the day-to-day variation in the number of battle events as recorded by the Armed Conflict Location & Event Data Project (ACLED) (in blue) as well as the share of countries where war-related tweets are trending (in orange). Attention was highest at the time of the invasion in February 2022 and the days of the Wagner Group rebellion in June 2023. Overall, the correlation between twitter trends and conflict intensity is positive but relatively weak.

Figure 4. Frequency of war-related hashtags and intensity of conflict.

Note: The chart shows the number of daily battle events in Ukraine as classified by ACLED on the left axis (in blue) and the share of countries where war-related topics were trending on the respective day on the right axis (in orange). The sources are ACLED’s Ukraine conflict monitor and data scraped from archive.twitter-trending.com

Attention also reacts to other major global events. Figure 5 compares the number of top-five trending hashtags related to the categories of interest in each country on two specific dates: February 24, 2022, the day of Russia’s full-scale invasion of Ukraine, and October 7, 2023, the day of a Hamas terror attack on Israel. On the day of the Russian invasion, the majority of countries in our sample exhibited the highest value. In contrast, on the day of the Hamas attack, related hashtags were trending almost nowhere outside Ukraine and Russia, indicating that global attention and engagement with this new ongoing crisis significantly overshadowed the focus on the situation in Ukraine. This shift in attention demonstrates how breaking news can capture the public’s interest and divert focus from ongoing crises, affecting the level of engagement on social media and potentially influencing the global response and discourse surrounding these events.

Figure 5. Map of prevalence of war-related hashtags on two different dates.

Note: The maps show the share of the top five highest volume topics on twitter related to Russia’s war on Ukraine. The map on the left shows 24/02/2022 – the day of Russia’s invasion. The map on the right shows 07/10/2023 – the day of a Hamas terror attack on Israel. Countries in white are not among the 62 countries in the dataset. The source are data scraped from archive.twitter-trending.com

While some events impact attention globally, others affect the salience of the conflict for a specific country. Figure 6 shows that people pay more attention to the war when there is a tangible connection to their own country. The panel on the left shows that war-related topics were more likely to trend in a country around the days where the country announced an aid package for Ukraine (military, financial or humanitarian). It shows an increasing trend in the preceding days and a peak on the day of the announcement. The panel on the right shows that war-related topics were more likely to trend in a country around the days of a visit from President Zelenskyj. This effect is large in magnitude but only lasts for around three days.

Figure 6. Likelihood of hashtags trending in relation to country-specific event.

Note: The charts show variation in the share of countries where at least one war-related topic was among the top five highest volume topics on days relative to a specific event. In the left chart, day 0 represents the day on which a country’s government announces an aid package for Ukraine. In the right chart, day 0 represents the day on which President Zelenskyj arrived in a country for an official visit. The source for these charts are: (i) the Kiel Institute’s Ukraine Support Tracker (Trebesch et al., 2023), (ii) Wikipedia’s list of official visits by President Zelenskyj and (iii) data scraped from archive.twitter-trending.com

While the events above act as drivers of attention, it is also interesting to consider what causes war-related topics to drop out of the top five trending topics. We distinguish between two reasons why war-related hashtags could stop trending: (i) a loss of interest that results in a reduction in the absolute number of related tweets (ii) the rise of other topics that displace war-related tweets from the top five. Figure 7 focuses on days where war-related topics dropped out and compares the volume of tweets on the last day where they were in the top five, to the threshold they would have had to surpass in order to make the top five on the subsequent day. In cases where the threshold is lower than the previously observed volume of tweets (a ratio of less than 1), the topic would have kept trending had it sustained its volumes, and one can conclude there was an absolute loss of interest. In cases where the ratio is greater than one, it is possible that the topic sustained its previous volume of tweets but was crowded-out by the rise of a new trending topic. Figure 7plots the histogram of this ratio. 73 percent of the cases are in the first category (loss of attention) and 27 percent are in the possible crowding out category. This provides further evidence to suggest that attention to the war on social media is typically fleeting.

Figure 7. Loss of attention vs crowding out.

Note: The sample are country-days where war-related topics were among the top five highest-volume topics but then dropped out of the top five the next day. The chart provides a histogram of the ratio of the threshold for making the top five on the subsequent day to the highest volume of tweets of a war-related topic. Values below 1 (in blue) indicate that the volume was above the next day’s threshold and the topic declined in absolute terms. Values above 1 (in orange) indicate that the volume was below the next day’s threshold. The source are data scraped from archive.twitter-trending.com

We also examine the content of discussions on the first day after war-related hashtags drop out of the top five. The word cloud in Figure 8 suggests that on such days, people primarily discuss entertainment topics like music and football.

Figure 8. Word cloud of hashtags trending on days war-related categories drop out.

Note: The figure provides a word cloud of trending topics on country-days where no war-related topic was among the top five highest volume topics, but at least one war-related topic had been in the top five on the previous day. The source are data scraped from archive.twitter-trending.com

Content and Context of War-Related Discourse

In addition to providing insight into the level of engagement, hashtag analysis can also reveal the content and context of popular discourse surrounding the war. By examining words trending on the same days as those from our 11 categories, we can gain a better understanding of the topics people are discussing and how the conversation varies across different regions. Figure 9 illustrates this through word clouds, showing the language used in NATO countries on the left and in countries that abstained or voted against the United Nations General Assembly Resolution ES-11/1 on the right. This resolution, dated March 2, 2022, condemned the brutal invasion of Ukraine and demanded that Russia immediately withdraw its forces and comply with international law.

This exercise allows us to compare the dominant themes and narratives in these two groups of countries and observe any differences in public perception and discourse regarding the conflict. The prevalence of cryptocurrency and NFT (non-fungible tokens) references in the word cloud on the right is suggestive of how economic interests and alternative financial systems could be relevant for the positions of countries that abstained or voted against the resolution, and how this might affect their involvement or response to the conflict. On the left, words like “NATO”, ”Biden”,  and ”Trump” clearly stand out, suggesting that these topics are central to the discourse on the war in NATO countries. This could indicate a focus on geopolitical alliances, international cooperation, and the role of key political figures in shaping the response to the conflict. Interestingly, “Putin” is very prominent in the left word cloud while “Russia” and “Russian” are more prominent on the right. This could indicate that Putin is seen and discussed as the primary antagonist in NATO countries.

Figure 9. Word cloud of hashtags in NATO countries vs Russia-friendly countries.

Note: These word clouds represent topics that trend on days where at least one war-related topic is trending in the respective country. The cloud on the left shows NATO countries. The cloud on the right shows countries that either abstained or voted against United Nations General Assembly Resolution ES-11/1. The source are data scraped from archive.twitter-trending.com

Conclusion

This brief uses X/Twitter trends as a barometer of public interest in Russia’s war in Ukraine. We show how attention fluctuates over time in response to developments in the conflict, to other breaking news, and to local events that make the conflict salient for a domestic audience. We also provide descriptive evidence on the variation across geographical regions and among different groups of countries. Additionally, we analyse the instance where Ukraine-related topics stop trending and find suggestive evidence that this is typically due to a gradual loss of interest rather than crowding out by new distracting trends.

Public attention and engagement drive policy in democratic countries, and the sustained support of its democratic allies is vital for Ukraine during this critical time. Understanding the patterns and influences of public attention is crucial for developing effective strategies to sustain engagement and support. This can be achieved for example by regularly highlighting the ongoing significance and bearing of Russia’s war against Ukraine, even as other events dominate the headlines. Emphasizing the impact of the conflict on individuals and communities, as well as its broader implications for international relations and global security, can help sustain public interest and engagement.

References

  • ACLED. Ukraine Conflict Monitor. https://acleddata.com/ukraine-conflict-monitor/
  • Trebesch, C., Antezza, A., Bushnell, K., Bomprezzi, P., Dyussimbinov, Y., Frank, A., Frank, P., Franz, L., Kharitonov, I., Kumar, B., Rebinskaya, E., Schade, C., Schramm, S., and Weiser, L. (2023). The Ukraine Support Tracker: Which countries help Ukraine and how? Kiel Working Paper, 2218, 1-75.
  • Twitter Trending Archive. Scraped on ##/12/2023. https://archive.twitter-trending.com/

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.

Insights and Research Shared at the 2023 FREE Network Retreat

FREE network retreat Image from the conference

The 2023 FREE Network Retreat, an annual face-to-face event for members of the FREE Network, gathered its representatives to share and exchange research ideas and to discuss its institutes’ respective work and joint efforts within the Network. An academic session highlighted multiple overarching areas of interest and opportunities for research collaboration and included a plenary session on topics ranging from theoretical underpinning of Vladimir Putin’s regime to climate change beliefs and to consumer behaviour in credit markets. A session addressing the respective institute’s work during the last year also demonstrated the importance and relevance of the FREE Network’s joint initiatives on gender, democracy and media, and climate change and environment: FROGEE, FROMDEE and FREECE. This brief gives a short outline of the plenary session and an overview of some further topics covered during the conference.  

The Academic Day

The Academic Day consisted partly of a plenary session and partly of an academic session. The academic session was outlined to demonstrate the wide spectrum of research interests within the network and to promote and highlight the opportunities for research collaboration. Designed as a series of poster sessions, each organized around a common research theme, it allowed for an exchange of ideas between presenting researchers and the audience while displaying the overlap of the various research interests across the institutes. At the same time, the poster session combined the broad range of topics within 10 overarching subjects (trade, gender, migration and education, public economics, energy, labor, political economy and development, macro, conflict, and theory and auctions).

The plenary session further illustrated the wide variety of topics the FREE Network researchers’ work on. During the plenary session, three distinguished presentations were held, summarized in what follows.

“Why Did Putin Invade Ukraine? – A Theory of Degenerate Autocracy”

Firstly, Konstantin Sonin, Professor at the University of Chicago Harris School of Public Policy, gave a presentation of his working paper (with Georgy Egorov, Northwestern University) in which the Russian full-scale invasion of Ukraine is explained through a theoretical framework on dictators’ decision-making in degenerate autocracies.

Sonin outlined how the beliefs about Ukraine in Kremlin, prior to the invasion, were factually wrong. For example, Kremlin believed that Ukraine, despite plenty of facts pointing in the opposite direction, lacked a stable government and had an incapable army. Further, it was believed that the US and Europe wouldn’t care about Ukraine and that Russian troops would be welcomed as liberators – the latter exemplified by the fact that Russia sent police and not the army during the first phase of the invasion. He also stressed that the decision to invade Ukraine is likely to have disastrous consequences for Vladimir Putin, his regime, and for Russia as a whole. This is, however, not the first example of a disastrous decision made by a leader of an autocratic regime, leading up to the question: What explains such choices that should not rationally have been made? And how can leaders make them in highly institutionalized environments where they are surrounded by councils and advisors who are supposed to possess the best expertise?

The model presented by Sonin assumes a leader in such highly institutionalized environment that wishes to stay in power and whose decisions are based on input from subordinates. The subordinates differ in level of their expertise and the leader thus chooses the quality of advice that he receives through his choice of subordinates.  In turn, while giving advice to the leader, the subordinate considers two factors: the vulnerability of the leader and their own prospects should the leader fall. In equilibrium there is a tradeoff as competent subordinates are also less loyal (since a more competent person might know when to switch alliances and have better prospects if the regime changes).

The leader also has access to repression as an instrument. Repression decreases his changes to be overthrown but raises the stakes for a potential future power struggle, as a leader with a history of repression is more likely to be repressed by his successor.

This interaction creates a feedback loop. If a dictator chooses repression, he feels more endangered, and he then chooses a more loyal subordinate who is less likely to deceive him for personal gain under a potential new regime. However, this leads to the appointment of less competent subordinates whereafter the information that flows to the leader becomes less and less reliable – as illustrated by Kremlin’s beliefs about Ukraine prior to the war.

There are three types of paths in equilibrium, Sonin explained; 1. “stable autocracy”, with leaders altering in power and choosing peaceful paths without repressions 2. “degenerate autocracy” – where the incumbent and opponent first replace each other peacefully and then slide into the repression-based change of power (until one of them dies and the story repeats), and 3. “consecutive degenerate autocracy” – where each power struggle is followed by repression.

Concluding his presentation, Sonin highlighted that in a degenerate autocracy such as Russia, individual decisions by the leader are rarely crucial due to the high level of institutionalization. However, as shown by the model, the leader is inevitably faced with a situation where he is surrounded by incompetent loyalists feeding him bad intel and setting him up to make disastrous decisions – most recently displayed in Vladimir Putin’s decision to invade Ukraine.

“Facing the Hard Truth: Evidence from Climate Change Ignorance”

Pamela Campa, Associate Professor at Stockholm Institute of Transition Economics, gave the conference’s second presentation, which detailed her work (with Ferenc Szucz, Stockholm University) on climate change skepticism.

Campa opened her talk with the current paradox regarding climate change, where, in the scientific community there is a strong consensus about the existence of climate change, but in society at large, skepticism is largely prevalent. This can be exemplified by one quarter of the US population not believing in global warming in 2023, and Europeans not believing in the fact that humans are the main driver of climate change.

According to Campa, the key question to answer is therefore “Why does ignorance about climate change persist among the public – in spite of the overwhelming evidence?”. One possible explanation may be a deficit in comprehension; people simply don’t understand the complexity of climate change and thus follow biased media and/ or politicians more or less sponsored by lobbyists. However, research have shown scientifical literacy to be quite uncorrelated with climate change denial, contradicting the above explanation. The second hypothesis, and of focus in the study, instead revolve around the concept of information avoidance. To test the hypothesis that people actively avoid climate change information, the authors key in on coal mining communities in the US having been exposed to negative shocks in the form of layoffs. These communities are of interest given their strong sense of identity and the fact that they are directly affected by the green transition. Arguably, a layoff shock would negatively affect not only their economy, but also pose a threat to their perceived identity. Given the context, it can thus be assumed that these communities to a larger extent would avoid information on climate change and information post-shock to restore the threatened identity.

The authors consider US counties experiencing mass layoff (more than 30 percent of mining jobs lost between 2014 and 2017) as treated counties, finding that in these counties, learning about climate change is 30 to 40 percent lower than in counties having experienced no mass layoffs. To account for the fact that the layoff itself may cause changes in learning, the authors also consider an instrument variable analysis in which gas prices are exploited as instrument for the layoffs – once again displaying the fact that people in affected communities believe climate change to be caused by humans to a lesser extent, when compared to counties in which no mass layoffs had occurred.

Interestingly, when controlling with other industries with somewhat similar characteristics (such as metal mining), the drop in climate change learning disappears, feeding in the notion of “identity-based information avoidance”.

The lack of support for and consensus among the public of the ongoing climate change and its drivers might pose a threat for the green transition as well as reduce personal effort to reduce the carbon footprint, Campa concluded.

“Consumer Credit with Over-Optimistic Borrowers”

In the plenary session’s last presentation, Igor Livshits, Economic Advisor and Economist at the Federal Reserve Bank of Philadelphia, presented his working paper (with Florian Exler, University of Vienna, James MacGee, Bank of Canada and Michèle Tertilt, Mannheimer University) on consumer credit and borrower’s behaviour.

There has been much debate on whether and how to regulate consumer credit products to limit misuse of credit. In 2009/2010 several initiatives and regulations (such as the 2009 Credit Card Accountability Responsibility and Disclosure Act) were introduced with the aim of protecting consumers and borrowers from arguments that sellers of credit products exploit lack of information and cognitive capacity of borrowers. There is however a lack of evaluation of such arguments and subsequent regulations, which Livshits explained to be the motivation behind the paper.

The paper differentiates between over-optimistic borrowers (behaviour borrowers) and rational borrowers (rationalists). While both types face the same risks, behaviour borrowers are more prone to shocks and are at the same time unaware of these worse risks (i.e., they believe they are rationalists). Focusing on these types of borrowers, the paper introduces a model in which the lenders endogenously price credit based on beliefs about the borrower type. Households decide whether to spend or save and if to file for bankruptcy in an environment in which they are faced with earning shocks and expense shocks.

In this structural model of unsecured lending and default, Livshits finds that behavioral borrowers’ “risky” behaviour negatively affects rationalists since both types are pooled together and, thus rationalists are overpaying to cover for the behaviour borrowers. A calibration of the model also suggests that behavioral borrowers borrow too much and file for bankruptcy too little and too late.

Livshits argued that the model does not provide evidence of the notion that borrowers need protection from lenders, but rather that borrowers need to be protected from themselves. In fact, had behaviour borrowers been made aware of the fact that they are overly optimistic about the actual state of their future incomes, they would borrow 15 percent less.

To address the increased risks behaviour borrowers take at the cost of rationalists, policies such as default made easier, taxation on borrowing, financial literacy efforts and score-dependent borrowing limits could all be considered. Such policies may lower debt and reduce bankruptcy filings but as they may also reduce welfare and exhibit scaling difficulties.

Updates from the Institutes

During the Retreat, the respective institutes shared the previous year’s work, and updates within the FREE Network’s three joint projects were also presented. These go under the acronyms of FROMDEE (Forum for Research on Media and Democracy in Eastern Europe), FREECE (Forum for Research on Eastern Europe; Climate and the Environment) and FROGEE (Forum for Research on Gender Economics in Eastern Europe), and address areas of great relevance in Eastern Europe and the Caucasus. Researchers from all FREE Network institutes work on these topics, with the most recent policy paper written in coordination by SITE, KSE and CenEA (with expert Maja Bosnic, Niras International Consulting). The policy paper focuses on the gender dimension of the reconstruction of Ukraine – putting emphasis on the necessity of gender budgeting principles throughout the various parts of reconstruction.  An upcoming joint research paper will consider the effects of gasoline price increase on household income across the Network’s countries, written under the FREECE umbrella.

The three themes of gender, media and democracy, and environment and climate are not only purely research topics within the institutes. They also reflect developments and challenges that the institutes to a various extent face in the respective contexts in which they operate. The work focusing on the reconstruction of Ukraine is an excellent example of an area that encompasses all three.

Another example of the relevance of the three themes features prominently in one of the institutes’ most tangible contribution to their respective societies: their education programs. Nataliia Shapoval, Vice President for Policy Research at Kyiv School of Economics (KSE), emphasized how KSE has – amid Russia’s war on Ukraine – managed to greatly expand. Over the past year, KSE has launched 8 new bachelor’s and master’s programs, some of which are directly targeted at ensuring postwar reconstruction competence. On a similar note, Lev Lvovskiy, Academic Director at the Belarusian Research and Outreach Center (BEROC) mentioned the likelihood of next year being able to offer students a bachelor’s program in economics and several business courses in Vilnius – BEROC’S new location. BEROC’s effort in providing quality education in economics to Belarus’ exile youth is considered a fundamental investment in the future of the country – providing a competent leading class capable of installing democracy and fair elections in Belarus once the current regime is gone. The emphasis on education was further highlighted by Salome Gelashvili, Practice Head, Agriculture & rural policy at the International School of Economics Policy Institute (ISET-PI) who not only mentioned the opening of a master’s program in Finance at ISET but also the fact that an increasing number of students who’ve recently graduated from PhD’s abroad are now returning to Georgia. Such investments into education are necessary to counter Russian propaganda in the region all three agreed, emphasizing the need to continually stem Russia’s negative influence in the region. This investment into education is also important to hinder countries from sliding away from democratic values – realized in Belarus and threatening in Georgia.

To further delve into the issues of democratic backsliding, a tendency that has been recently observed not only in the region but also more widely across the globe, FROMDEE will organize an academic conference in Stockholm on October 13th, 2023.

Concluding Remarks

The 2023 FREE Network Retreat provided a great opportunity for the Networks’ participants to jointly take part of new research and to share experiences, opportunities, and knowledge amongst each other. The Retreat also served as reminder of the importance of continuously supporting economic and democratic development, through research, policy work, and networking, in Eastern Europe and the Caucasus.

List of Presenters

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 Coverage and Pandemic Behaviour: Evidence from Sweden

20230412 Media Coverage and Pandemic Behaviour Image 01

Individual behaviour can often have wider societal consequences and it is important to understand how to affect positive behavioural change. In this policy brief, we document the ability of the media to increase pro-social behaviour during a public health crisis. In Garz and Zhuang (2022), we collect a unique dataset of 200,000 newspaper articles about the Covid-19 pandemic from Sweden – one of the few countries that did not impose mandatory lockdowns or curfews but largely relied on voluntary social distancing. We show that mentions of Covid-19 significantly lowered the number of visits to workplaces and retail and recreation areas, while increasing the duration of stays in residential locations. The impacts are largest when Covid-19 news stories were more locally relevant, more visible and contained simple and explicit public health advice. These results have wider implications for the design of public communications and the value of the local news media.

The Covid-19 pandemic had devastating health and economic consequences for the entire world (see, for instance, the COVID19 | FREE Network Project for more information on the pandemic experience of the FREE network countries). To stop the spread of the virus, many governments imposed curfews or lockdowns. These mandatory restrictions on people’s movements were and remain controversial. In several countries, these public health policies fuelled conspiracy theories and led to protests and refusals to adopt other protective measures, such as wearing face masks. The policies have also been criticised for placing the burden disproportionately on the poorest members of society, many of whom lost their livelihoods due to the restrictions.

Sweden is one of few countries in the world that did not to impose any lockdowns or curfews. Instead, the Swedish strategy relied on voluntary compliance with public health recommendations aimed at reducing mobility and encouraging social distancing (for FREE policy briefs featuring in-depth discussions of the Swedish pandemic response and experience, see Becker et al., 2020; Hauser, 2020a; Hauser, 2020b; Campa, Roine and Strömberg, 2021; Berlin, 2020.) This strategy was by and large effective, with Sweden seeing similar falls in average mobility in 2020 compared to its Scandinavian neighbours which however imposed strict lockdowns.

What are the drivers of voluntary compliance? A recent SITE working paper “Media Coverage and Pandemic Behaviour: Evidence from Sweden” by Marcel Garz (Jönköping University) and Maiting Zhuang (SITE, Stockholm School of Economics) sheds light on this question. We show that news coverage of the pandemic could have played an important role in shaping public opinion, social norms and ultimately individuals’ health-related actions in Sweden. In this brief, we summarise our approach and findings, and discuss important policy insights.

Data

We collect close to 200,000 newspaper articles about Covid-19 from Swedish newspapers during 2020, which represents almost all newspaper coverage of the pandemic during that year in Sweden. Newspapers remain a major source of information in Sweden, where close to two thirds of the population are regular newspaper readers. We analyse the full texts of articles to identify which aspects of news coverage have the largest impact on behaviour. Figure 1 shows Covid-19 coverage in Swedish newspapers in 2020.

Figure 1. Newspaper coverage of Covid-19

Source: Retriever and authors’ calculations. This graph shows the number of articles including “coronavirus* OR covid*” on a given day, weighted by newspaper circulation. The grey solid lines are local polynomial smooths with 95 percent confidence intervals.

As our main outcome variable, we use Google Community Mobility Report data – anonymous data on the number (or duration) of visits to different types of locations (such as retail and recreation, workplaces, or residential) within a municipality.

Figure 2. Workplace and residential mobility

Source: Google and authors’ calculations. These graphs show the daily percentage change in mobility compared to median values between January 3 to February 6, 2020. The grey solid lines are local polynomial smooths with 95 percent confidence intervals.

Method

We want to study whether reading news concerning Covid-19 affected people’s decisions to reduce their own mobility in Sweden. However, determining the effect of the media on behaviour is difficult for several reasons. First, people choose what media to consume and most often tend towards media outlets that confirm their existing views. Second, there could be factors that determine both media coverage and behaviour.

In our paper, we address the first concern by carefully choosing our data. If we found that people who read more about Covid-19 in newspapers are also more likely to reduce their mobility, this could be driven by many different individual factors, such as how worried a person is about the pandemic. Instead, we measure exposure to Covid-19 news using data on the number of subscriptions every newspaper sells in each municipality. As newspaper subscriptions are annual and were decided before the pandemic, this measure of Covid-19 news exposure should not be correlated with individuals’ beliefs about the pandemic.

The second concern is more relevant, as it is likely that media coverage and individual mobility are both driven by the spread of the pandemic. To alleviate this concern, we adopt several strategies. The first is that we take into account local pandemic severity using a range of different measures, including excess deaths, infections, Covid-19 deaths at different geographic units and with different time lags.

Our second strategy relies on the fact that Swedish newspapers typically circulate in multiple municipalities, but are more likely to respond to events in municipalities where more of their subscribers live. We use an instrumental variable (IV) approach where we use the circulation-weighted excess mortality in a newspaper’s distribution area as the IV. This IV is a strong predictor for the amount of Covid-19 coverage by that newspaper but should not affect mobility in a target municipality – conditional   on the pandemic severity in the municipality itself. We also show that our results hold in a sample of “peripheral” municipalities, that is municipalities that only form a small percentage of any newspaper’s subscriber base, but where these newspapers are nonetheless major sources of information.

Results

We find evidence that media coverage of the pandemic increases compliance with the main public health recommendation to work from home in Sweden. More newspaper articles about Covid-19 on a given day and in a given municipality is associated with increased residential mobility and lower workplace mobility (Table 1), as well as fewer visits to retail areas and recreational facilities. This pattern remains even when controlling for the severity of the pandemic at the local level. The results are robust using different methods and data.

Table 1. Effect of Covid-19 coverage on workplace and residential mobility

Source: Retriever, Google, SMHI, MPRT and authors’ calculations. Observations are at the municipality-day level. The dependent variables are percentage changes in Google mobility for workplaces (columns 1 and 3) and residential areas (columns 2 and 4) relative to the baseline. The main regressor is the circulation-weighted number of Covid-19 articles. Columns 3 and 4 show IV results using the interaction between weighted excess deaths across a newspaper’s circulation area and newspaper exposure as the IV. Local pandemic severity is the latest available excess deaths figure in the municipality at the time of newspaper publication. All regressions control for municipality and day fixed effects. Standard errors in parentheses are clustered by municipality. Significance at the 1%, 5% and 10% is denoted by ***, ** and *, respectively.

The type of coverage matters. Behaviour responds the most to personally relevant news that is easy to understand and visible. In particular, Covid-19 articles which explicitly mention the affected municipality have a larger impact on mobility than, for instance, articles that only relate to developments abroad. There are larger impacts of more factual compared to more subjective reporting. Articles that contain direct and explicit public health advice have a large impact on individual behaviour. In contrast, articles that mention medical experts have a smaller impact on individual behaviour – likely due to the complexity of the language used.

We also find a greater impact on individual behaviour in response to more visible Covid-19 stories, such as articles on the front page or articles whose headlines mention the pandemic. These results are consistent with media coverage not just increasing the salience of the pandemic and reminding individuals to follow official guidelines, but also providing relevant information. Despite fears that the large amount of press coverage could lead to individuals avoiding news about the pandemic, we find little evidence for media fatigue except at very high levels of coverage.

Conclusion

In Garz and Zhuang (2022), we study the effects of media coverage on individual behaviour during a public health crisis. We focus on Sweden, a country that did not impose any lockdowns or curfews during the Covid-19 pandemic and where newspapers remain an important source of information. Using close to the universe of all Swedish newspaper articles about Covid-19 in 2020, we find that media coverage of the pandemic encouraged people to stay at home. The effect is largest when news stories are of local relevance and contain explicit public health advice. These results have important implications for the design of future public communication strategies that aim to foster behavioural change.

We also find little evidence of media fatigue or a preference of opinion pieces relative to factual reporting when it comes to Covid-19 in Sweden. While there has been much discussion about misinformation and media bias during the pandemic, our paper shows a positive effect of the local news media in terms of encouraging voluntary adherence to public health measures.

More broadly, our study adds an important dimension to the policy discussion about the decline of local news, beyond local political accountability and community participation. We find that local news remains an important source of local information, and that personally relevant information is more important for behavioural change. A lack of trusted local media could adversely affect compliance with government recommendations during a crisis, as well as a range of other campaigns, such as those encouraging the take-up of vaccines or adoption of more environmentally friendly behaviours.

References

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.

From East to West: A Paper Curtain in Swedish Foreign News Coverage?

Selective Focus Photography of Magazines representing Media Freedom Eastern Europe

How much a country is talked about in the media can determine its place in the public debate. In this brief, we collect data on the mentions of Eastern and Western European countries in the main Swedish newspapers over the past decades. We find consistently more coverage devoted to Western compared to Eastern Europe in the Swedish press. We investigate several factors that could explain this pattern. We find that while Eastern European countries are on average not more geographically distant from Sweden, Sweden tends to have closer trade links with Western European countries. Sweden is more culturally similar to the average Western European country in terms of language, religion and attitudes, cultural values and social norms. Trade relations and cultural proximity are associated with higher media coverage.

The media plays a vital role in modern societies by keeping the public informed and policymakers accountable. Whether and how events are covered by the news determines their relevance in the public debate. There is ample empirical evidence on the agenda-setting power of the news media. For example, Snyder and Strömberg (2010) show that local press coverage affects how informed US voters are about their representatives and in turn how much their politicians work in the interest of their constituencies. Eisensee and Strömberg (2007) find that news coverage affects how much disaster relief the US sends to foreign countries.

In this brief, we study the amount of news coverage devoted to European countries in the Swedish press. We document a systematic difference between Western and Eastern Europe and explore underlying factors that could be important in explaining this East-West divide.

The East-West Divide

We choose the four most widely read Swedish newspapers (Aftonbladet, Expressen, Dagens Nyheter, and Svenska Dagbladet) and use the newspaper database Retriever Research Media Archive to obtain statistics on the number of mentions of each country between 1995 and 2021. A country mention is an article in which the name of a country appears. Since two or more countries can be named in the same article, the total number of mentions does not correspond to the number of articles. As a percentage of all articles published by the four newspapers in 2021, roughly 20% mention at least one of these countries. While this simple measure of news coverage can be informative, it does not take into account many other aspects of a country’s prominence in the news, such as the length of articles, where articles appear, the tone of coverage, etc.

Figure 1 plots the sum of annual number of mentions by region over time. We see a clear difference in the amount of coverage devoted to Eastern and Western European countries. Over the entire time period, the 21 Western European countries were mentioned on average 2.7 times more than the 22 Eastern European countries.

While there does not appear to be a trend in relative coverage, there is considerable variation from year to year. The year when the relative difference in the number of mentions is smallest is 2014. The two most mentioned Eastern European countries in that year were Russia and Ukraine. Coverage likely increased due to the Crimean Crisis, when Russia invaded and annexed the Crimean Peninsula in Southern Ukraine. The relative difference was also low in 2008, coinciding with the Russo-Georgian war in August. In that year, other newsworthy events, such as the Global Financial Crisis or the UEFA European Football Championship, have a more ambiguous effect on relative media coverage.

Figure 1. Country mentions in Swedish newspapers

Note: Countries included in Eastern Europe: Albania, Armenia, Azerbaijan, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic, Estonia, Georgia, Hungary, Latvia, Lithuania, Moldova, Montenegro, North Macedonia, Poland, Russia, Serbia, Slovakia, Slovenia, Ukraine. Countries included in Western Europe: Andorra, Austria, Belgium, Denmark, Finland, France, Germany, Iceland, Ireland, Italy, Liechtenstein, Luxembourg, Malta, Monaco, Netherlands, Norway, Portugal, San Marino, Spain, Switzerland, United Kingdom.

What Explains This Discrepancy Between East and West?

There are a number of potential reasons why some countries systematically receive more attention in the press. In this section, we correlate the mean annual mentions of each country between 2019 and 2021 with different aspects of that country’s relationship with Sweden.

Distance and Population

Figure 2 shows how news coverage of a country depends on its geographic distance to Sweden and its population size. Overall, the further a country is from Sweden, the less that country is covered in the Swedish press. On average, Eastern European countries (in yellow) are covered less than Western European countries (in blue), for a given distance to Sweden. For example, Poland and Germany are both around 1000km away from Sweden, but Germany is mentioned almost twice as often in the Swedish press. As we measure the distance between the most populous city of each country and Stockholm, some of this difference in coverage is driven by the fact that countries sharing a border with Sweden receive extensive coverage. For instance, Denmark, Finland, and Norway are on average covered more than six times as much as Latvia.

Population also plays a role, that is, larger countries (e.g., Germany, Russia, Spain, and Poland) receive more coverage than smaller countries (e.g., Lithuania, Ireland, and Estonia). As Eastern European countries have on average smaller populations than Western European countries, population can partly explain the East-West difference in news coverage. One counterexample is Russia, which has more than twice as many people as France or the UK, but receives less coverage in the Swedish press.

Figure 2. Geographical distance and population

Note: Geodesic distances are calculated between the latitudes and longitudes of the most populous city of each country and Stockholm. Marker sizes are weighted by population averaged over 2019-2021, and fitted line is unweighted. Source: CEPII’s GeoDist dataset (Mayer and Zignago, 2006) and the World Bank. See Figure 1 for a list of countries included.

Trade and GDP

Figure 3 shows that Sweden’s economic relationship with a country affects how much the country features in Swedish news. We find a strong positive correlation of 0.8 between a country’s total trade volume with Sweden and country mentions in Swedish newspapers. As Sweden’s largest trading partners tend to be in Western Europe, this partly explains the relative coverage of East and West. Another factor is the overall size of a country’s economy (as measured by its GDP). Swedish newspapers more commonly mention countries with higher GDP, and these are more likely to be in Western than Eastern Europe.

 Figure 3. Trade and GDP

Note: Trade data are from 2019. Marker sizes are weighted by national GDP, and fitted line is unweighted. GDP figures are averaged over 2019-2021 and measured in current prices, PPP adjusted, international dollars. Source: The World Bank’s WITS database and the IMF World Economic Outlook, October 2021. See Figure 1 for a list of included countries.

Culture

There is a large literature documenting the link between cultural factors and the economic relationship between nations. For instance, studies show that similarities in ancestry, language, religion, norms and values can influence bilateral trade (Melitz, 2008; Guiso et al., 2009) and the diffusion of technology (Spolaore et al., 2009). In this section, we show how the amount of press coverage correlates with differences in language, religion, and values and norms using cultural distance data from Spolaore and Wacziarg (2016).

Figure 4.a shows that Swedish newspapers are more prone to cover countries whose languages are similar to Swedish. The language similarity measure originally developed by Fearon (2003) is based on the prevalence of languages within a country and distance between languages. The distance measure is calculated using linguistic trees provided in Ethnologue. It ranges from 0 (close) to 1 (distant) and reflects the expected number of common linguistic nodes between two randomly chosen individuals from each country and takes into account that countries can be linguistically heterogeneous (for more details, see Fearon 2003). Norway and Denmark are linguistically closest to Sweden, however, these are also two neighboring countries with which Sweden conducts extensive trade. On average, Eastern European countries are more linguistically distant from Sweden, although some Western European countries (such as France and Spain) are as linguistically distant from Sweden as many of the Eastern European countries and receive considerably more press coverage.

The religious distance measure by Spolaore and Wacziarg (2016) is calculated analogously to the linguistic distance measurement. It is based on the prevalence of different religions within a country and the distance between religions. Figure 4.b shows that countries that are religiously different from Sweden receive less coverage in the Swedish media. With the exception of the three Scandinavian countries, Eastern and Western European countries have similar levels of religious distance to Sweden. Based solely on this metric, the Swedish press mentions Eastern European countries less (and Western European countries more) than their religious distance to Sweden would predict.

Figure 4.c shows an index of a country’s cultural proximity to Sweden, that is, its distance in terms of cultural values, attitudes and norms based on average responses to the World Value Surveys from 1981 to 2010 (see Spolaore and Wacziarg, 2016). This cultural proximity index aggregates the Euclidian distances in survey responses between each country and Sweden. The index is standardized so that 0 shows the average country’s cultural distance to Sweden and negative (positive) values indicate above (below) average cultural similarity. Western European countries are significantly closer to Sweden than Eastern European countries based on this measure. As Swedish press coverage is on average declining in a country’s cultural distance to Sweden, this difference in country’s values and attitudes can explain some of the East-West difference in media coverage.

 Figure 4. Cultural distance

Panel a. Linguistic distance

Note: We use the indicator of tree-based weighted linguistic distance from Spolaore and Wacziarg (2016) and originally developed in Fearon (2003). This measure is an estimate of the expected or weighted number of common linguistic nodes between two randomly chosen individuals from each country. The data on language prevalence is compiled from a number of different sources and assembled in Fearon (2003). Countries included in Eastern Europe: Albania, Armenia, Azerbaijan, Belarus, Bulgaria, Croatia, Czech Republic, Estonia, Georgia, Hungary, Latvia, Lithuania, Moldova, Poland, Russia, Slovakia, Slovenia, Ukraine. Countries included in Western Europe: Austria, Belgium, Denmark, Finland, France, Germany (average between East and West Germany), Ireland, Italy, Netherlands, Norway, Portugal, Spain, Switzerland, United Kingdom

Panel b. Religious distance

Note: We use the tree-based weighted religious distance from Spolaore and Wacziarg (2016). This measure is an estimate of the expected distance between the religions of two randomly chosen individuals from each country. See Figure 4.a for a list of included countries.

 Panel c. Distance in cultural values, attitudes, and norms

Note: We use the distance in cultural norms and values from Spolaore and Wacziarg (2016).  This measure is based on all value-related questions from the World Values Survey Integrated Questionnaire from 1981–2010. The mean distance across countries is standardized to zero. See Figure 4.a for a list of countries included.

Conclusion

As the public and policymakers primarily receive information from the mass media, news coverage can have profound effects on public debate and policy decisions. Using data on the content of the four most widely read Swedish newspapers over the past decades, we measure how much the Swedish press covers Eastern and Western European countries. We find that over the past 25 years, there have been 2.7 times more mentions of Western than Eastern European countries. We find that the Swedish press is more likely to mention countries that are geographically closer, more populous, have a larger GDP and more trade with Sweden. Cultural proximity (as measured by language, religion and values, attitudes and social norms) also correlates with higher coverage. These factors are of course not independent from each other. For instance, the other Scandinavian countries with whom Sweden shares a border and a history, are culturally similar to Sweden and some of Sweden’s most important trading partners. They are also some of the countries that are most covered by the Swedish press. Some of these factors, such as sharing similar values, appear to explain the gap in coverage between East and West, while others, such as geographic distance, do not. More recently, concerns over energy security in the EU (see e.g., Le Coq and Paltseva, 2022) and the rise in military tension between Russia and Ukraine illustrate how developments in Eastern Europe can directly affect life here in Sweden. Perhaps it is time for Sweden to pay more attention to her eastern neighbours?

References

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.

Introduction

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).

Conclusion

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.

References

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.

Conclusion

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.

References

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.

Independent Media and Contemporary Military Doctrines

Governments often take unpopular measures. To minimize the political cost of such measures policy makers may strategically time them to coincide with other newsworthy events, which distract the media and the public. We test this hypothesis using data on the recurrent Israeli-Palestinian conflict. We show that Israeli attacks are more likely to be carried out when the U.S. news are expected to be dominated by important (non-Israel-related) events on the following day. In contrast, we find no evidence of strategic timing for Palestinian attacks.

The role of media in today’s conflicts is enormous. Parties to conflicts use propaganda in state-sponsored media and enroll state-sponsored trolls in social media to gain domestic public support for their military campaigns and, more generally, to raise own popularity. Involvement of Russia in Syria and Eastern Ukraine and its coverage on Russia-sponsored TV is a forceful illustration of this. Some most devastating conflicts used state media to enroll paramilitary. For example, Yanagizawa-Drott (2014) estimated that 51,000 perpetrators in Rwandan genocide were persuaded to participate in mass killings by RTLM radio.

Not all the media are under control of parties involved in conflicts. What is the role of independent media during conflicts? It is one thing to use the dependent media to portray one’s participation in conflict in a slanted manner; it is another to change one’s military strategy in order to improve one’s image in the independent media. Do military choose the timing and the weapon for their offences depending on the expectation of how their actions will be portrayed by the independent media? A statement on June 4, 2002, by Major General Moshe Ya’alon, then the Israel Defense Forces (IDF) chief of staff designate and until recently the defense minister of Israel, strongly suggests this is the case for the Israeli-Palestinian conflict. Mr. Ya’alon said: “This is first and foremost a war of ideology, and as such the media factor, the psychological impact of our actions, is critical. If we understand that a photograph of a tank speaks against us on CNN, we can take this into account in our decision as to whether or not to send in the tank. We schedule helicopter operations for after dark so they cannot be photographed easily. … Such considerations are already second nature to us. Officers … must understand that there are strategic media considerations. The tension between the need to destroy a particular building or to use a tank or helicopter, and the manner, in which the world perceives these actions, can affect the ultimate success or failure of the campaign. Even if we triumph in battle, we can lose in the media and consequently on the ideological plane.”

Our recent paper “Attack When the World Is Not Watching? U.S. News and the Israeli-Palestinian Conflict” (Durante and Zhuravskaya, 2017) forthcoming in the Journal of Political Economy investigates how Israeli military changes the planning of its operations in Gaza and the West Bank in the face of coverage by US media. In particular, we test whether Israeli authorities choose the timing of their attacks strategically to coincide with other newsworthy events so as to minimize the negative impact of their actions on U.S. public opinion by avoiding U.S. media coverage of their military operations, especially when they might lead to civilian casualties.

Methodology

We compile a list of fully exogenous events from forward-looking political and sports calendars in the U.S. between 2001 and 2011 and verify which of these events actually dominate US TV news, leaving little or no time to coverage of other events. Then, we compare the timing of these events to the timing of Israeli attacks on a daily basis.

We also use another, more continuous measure of whether the U.S. media and the public are distracted by other important events, namely the length of top three non-conflict-related news stories during evening news on three U.S. TV networks, where the evening newscasts are limited to 30 minutes, namely ABC, CBS, and NBC. As Eisensee and Stromberg (2007) point out, due to the competition between networks for audience, we can measure the importance of newsworthy events featured on the evening broadcasts because more important stories appear before less important stories, and they are longer.

Results

Timing of Israeli attacks and their coverage in US media

We find that both the incidence and the severity of Israeli attacks increased sharply when U.S. news were dominated by other events, such as US primaries and caucuses, general elections, and Presidential inaugurations. The probability that Israel carried out an attack against Palestinians rose to 53.2% one day before these important U.S. events from 38.7% on days that did not coincide with these events (over our observation period of 11 years, which includes heavy fighting during the Second Intifada). Figure 1 illustrates this finding. Attacks which coincide with the major political and sports events are also more deadly; as a consequence, the number of victims of Israeli attacks per day is 1.51 times higher during the days that coincide with major political and sports events compared to days that do not coincide with major events.

Figure 1. IDF attacks and exogenous predictable newsworthy events in the U.S.

Source: Durante and Zhuravskaya, 2017.

Using another measure, the length of top three non-conflict-related news stories during evening news on three U.S. TV networks, we also find that Israeli attacks are significantly more likely to occur and are more deadly when top three non-conflict-related news are longer on the following day.

Does it matter which military operation?

As some military operations are more costly to postpone than others, one should expect that only attacks that are less costly to more be strategically timed to other important events. This is exactly what we find: the timing of special targeted-killing operations, which are considered as extremely urgent by IDF, is not related to U.S. news cycle. In addition, one should expect military operations to be timed to other newsworthy events only when they are likely to generate negative publicity. As negative publicity about the conflict is mainly associated with civilian casualties, and civilian casualties are more likely when the operations are executed with heavy weapons, we find that the relationship between occurrence and severity of Israeli attacks and U.S. newsworthy events on the following day holds only for operations that involve the use of heavy weapons. We also check that the attacks are only timed to predictable newsworthy events.

Why tomorrow’s coverage matters more?

Israeli attacks get news coverage in U.S. media both on the day of the attack and one day later. Why, then, Israel times its attacks to news pressure on the following day rather than on the same day? To answer this question, we analyzed the content of news broadcasts and found that the type of coverage of Israeli attacks differs substantially between same-day and next-day reports. While the same-day and next-day news stories are equally likely to report information on the number of victims, news stories that appear on the day after the attack are much more likely to present personal stories of civilian victims and include interviews with their relatives or friends. Furthermore, next-day coverage is significantly more likely to include emotionally charged visuals of burial processions and scenes of mourning. Anecdotal evidence suggests that it is both easier and safer for a foreign journalist to get details of the story on the next day; and that the next day affords an opportunity to produce emotionally charged videos of funerals. Figure 2 illustrates these findings.

Figure 2. Comparison of the content of news casts about attacks that aired on the same day as an attack and on the day following the attack.

Source: Here you can write notes to the figure, graph or table. Do not forget to state the source of the figure, graph or table.

Since people react more strongly to personal stories than to statistics and facts, and since information transmitted only through words is less likely to be retained than information accompanied by images, it is not surprising that Israel times its attacks to predictable international newsworthy events expected on the following day, as the next-day news stories are more damaging to Israel’s public image.

Conclusion

These results have broader implications. Policy makers in other policy domains and other countries may also strategically manipulate the timing of their unpopular actions to coincide with other important events that distract the mass media and the public. Examples of unpopular policies characterized by suspicious timing abound: Silvio Berlusconi’s government passed an emergency decree that freed hundreds of corrupt politicians on July 13, 1994, the day Italy qualified for the FIFA World Cup final. Russian troops stormed into Georgia on August 8, 2008, the opening day of the Beijing Summer Olympics. Political spin-doctors often release potentially harmful information in tandem with other important events. This is exemplified by a notorious statement from the former UK Labour Party’s spin doctor, Jo Moore, who, in a leaked memo sent to her superiors on the afternoon of 9/11, said that it was “a very good day to get out anything we want to bury” (see http://www.telegraph.co.uk/news/uknews/1358985/Sept-11-a-good-day-to-bury-bad-news.html (accessed on July 7, 2015) and http://www.theguardian.com/politics/2001/oct/10/uk.Whitehall (accessed on July 7, 2015)).

Overall, policy makers’ strategic behavior may undermine the effectiveness of mass media as a watchdog, thus reducing citizens’ ability to keep public officials accountable

References

  • Durante, Ruben; and Ekaterina Zhuravskaya, 2017. “Attack When the World Is Not Watching? U.S. News and the Israeli-Palestinian Conflict”, Journal of Political Economy (forthcoming)
  • Eisensee, Thomas; and David Stromberg, 2007. “News Droughts, News Floods, and U.S. Disaster Relief,” Quarterly Journal of Economics, 05, 122 (2), 693–728.
  • Nevo, Baruch; and Shur Yael, 2003. The IDF and the press during hostilities, Jerusalem: The Jerusalem Democracy Institute, pp. 84-85, available at http://en.idi.org.il/media/1431355/IDFPress.pdf, accessed on May 18, 2016.
  • Yanagizawa-Drott, David, 2014. “Propaganda and Conflict: Evidence from the Rwandan Genocide,” Quarterly Journal of Economics, 129(4), pp.1947-1994.

Save

Save