INTRODUCTION

May 3, 2024 marks the 31st anniversary of World Press Freedom Day which serves as a “reminder to governments of the need to respect their commitment to press freedom… [as well as] a day of reflection among media professionals about issues of press freedom and professional ethics.”

To study where we stand in these areas, CNTI examined global trends across multiple measures over the last 31 years. The findings reveal worrying trends in global press freedom and safety. While protections from media censorship and harassment of journalists grew, on average, between 1993 and the early 2010s, these scores have fallen back down to below their 1993 levels.

What’s more, these recent declines cut across all four government regime types: liberal democracies, electoral democracies, electoral autocracies and closed autocracies. These findings support other research showing downward trends in press freedoms, including within government policies related to news information.

For this analysis, CNTI assessed global press freedoms using data from the Varieties of Democracy Institute (V-Dem) which has country-level metrics dating back to 1789. Two measures were of primary interest: (1) government censorship of print or broadcast media and (2) harassment of journalists. These metrics are a part of CNTI’s searchable dataset of 15 measures across 179 countries which also provides links to V-Dem’s full methodology and statistical approach. (The data examined below use the linearized transformation (i.e., _osp) variables which form an interval scale from 0 to 4.) 

An outline of the two measures is presented here:

Measuring Government Censorship

V-Dem’s measure of government censorship asks the question, “Does the government directly or indirectly attempt to censor the print or broadcast media?” (p. 207) with the following five response options:

• 0: Attempts to censor are direct and routine.
• 1: Attempts to censor are indirect but nevertheless routine.
• 2: Attempts to censor are direct but limited to especially sensitive issues. 
• 3: Attempts to censor are indirect and limited to especially sensitive issues.
• 4: The government rarely attempts to censor major media in any way, and when such exceptional attempts are discovered, the responsible officials are usually punished.

Measuring Harassment of Journalists

V-Dem’s measure of journalist harassment asks, “Are individual journalists harassed — i.e., threatened with libel, arrested, imprisoned, beaten, or killed — by governmental or powerful nongovernmental actors while engaged in legitimate journalistic activities?” (p. 209) with the following five response options:

• 0: No journalists dare to engage in journalistic activities that would offend powerful actors because harassment or worse would be certain to occur.
• 1: Some journalists occasionally offend powerful actors but they are almost always harassed or worse and eventually are forced to stop. 
• 2: Some journalists who offend powerful actors are forced to stop but others manage to continue practicing journalism freely for long periods of time.
• 3: It is rare for any journalist to be harassed for offending powerful actors, and if this were to happen, those responsible for harassment would be identified and punished.
• 4: Journalists are never harassed by governmental or powerful nongovernmental actors while engaged in legitimate journalistic activities.

Both metrics use lower scores to denote greater harm to press freedom and journalists while higher scores represent greater safety and protection. It is also important to note that the measure of government censorship specifies print and/or broadcast media and that the measure of journalist harassment does not capture harassment from the public. 

Global Trends in Media Censorship and Journalist Harassment Have Fallen Back Down to 1990s Levels

How have media censorship and journalist harassment changed over the last 31 years? The figure below shows the yearly global average values for government censorship of print and broadcast media and for journalist harassment from the roughly 179 counties in the V-Dem dataset from 1993 to 2023. (The number of countries classified by V-Dem varies slightly year-to-year. See the Appendix for a yearly breakdown.) Both of these metrics follow similar patterns (the correlation is r = 0.96). 

V-Dem’s measures (based on a scale from 0 to 4 with 0 being the worst score for press freedoms and 4 being best) show protection from both government censorship and harassment of journalists increased overall from 1993 through the early 2010s. After peaking in 2012, however, both of these measures decreased (down 11% and 9%, respectively), landing below their starting values in 1993. As such, global averages for both government censorship and journalist harassment were slightly lower (i.e., worse) in 2023 than in 1993. 

Government censorship of print or broadcast media had a global average of 2.16 in 2023, which in V-Dem’s definition means that “[a]ttempts to censor are direct but limited to especially sensitive issues.” The 2023 global average for harassment of journalists of 1.99 signals that “[s]ome journalists who offend powerful actors are forced to stop but others manage to continue practicing journalism freely for long periods of time.” 

There is certainly some variation by country, but the averages are a helpful way to understand the general state of our global society. Exploration of country-level information is available in CNTI’s searchable dataset. Below we offer an additional breakdown by regime type.

Government Regime Types and Press Freedoms

Another important consideration is the relationship between government structure and press freedoms. V-Dem provides a classification of countries’ political regimes with four categories: (1) liberal democracy, (2) electoral democracy, (3) electoral autocracy and (4) closed autocracy. The table below presents the breakdown of these categories using V-Dem’s most recent report for 2023 which classified 179 countries. Democracies constituted about 51% of governments while autocracies made up about 49% of governments. 

Liberal democracies experienced a 3.5% decrease in their scores from 2012 to 2023. Electoral democracies declined 1.6% during the same period. Electoral autocracies decreased 11%. Closed autocracies, which also consistently received the lowest (i.e., worse) baseline scores for journalist harassment, decreased 9% from 2012 to 2023. 

It is important to note that each regime type has different starting points and vertical ranges in the figure above. As such, even with the 3.5% decline, liberal democracies still have the highest average scores in 2023 compared to the other three regime types. Similarly, electoral autocracies performed slightly better, on average, than closed autocracies. 

Conclusion

This analysis sheds important light on the downward trend over the last decade for press freedoms around the world. And while scores have fallen more in some regime types than others, no type has, on average, strengthened their scores since 2012. 

The numeric size of the declines above, which range between 0.04 and 0.19 units, are not massive in part because the V-Dem measure is on a 5-point scale. However, even these small decreases correspond to significant percentage changes in these measures of approximately 2-11%, depending on regime type. The steady declines over the last decade are important for understanding the backsliding trajectory of press freedoms globally.

This research brief follows several other recent and ongoing projects CNTI is conducting. In January 2024, CNTI released a “fake news” report which found that proposed bills or existing laws on “fake news” and mis- and disinformation from 2020 to 2023 created opportunities for governments to curtail press freedoms. An issue primer on online abuse and harassment of journalists is also in development and will be released soon. For more information, please visit innovating.news.

APPENDIX

Note: The Center for News, Technology & Innovation (CNTI) has an aggregated data resource of metrics from sources including V-Dem, Reporters Without Borders, Committee to Protect Journalists, the World Bank and the World Justice Program among several others. Combining the metrics from these resources helps to understand global patterns. Given that these data only show the most recent year, the findings above are supplemented with V-Dem’s time series data. 

The code for this project may be found on CNTI’s GitHub here. 

About CNTI

The Center for News, Technology & Innovation (CNTI), an independent global policy research center, seeks to encourage independent, sustainable media, maintain an open internet and foster informed public policy conversations.

The Center for News, Technology & Innovation is a project of the Foundation for Technology, News & Public Affairs.