2020

16TH

JUNE

TIME)

(BRITISH

00.01

ReutersEMBARGOED: Institute Digital News Report 2020 2020

16TH

JUNE

TIME)

(BRITISH

00.01

EMBARGOED: Reuters Institute Digital News Report 2020

Nic Newman with Richard Fletcher, Anne Schulz, Simge Andı, and Rasmus Kleis Nielsen 2020

16TH

JUNE

TIME)

(BRITISH

00.01 Supported by

EMBARGOED:

Surveyed by

© Reuters Institute for the Study of Journalism Reuters Institute for the Study of Journalism / Digital News Report 2020 4

Contents

2020 Foreword by Rasmus Kleis Nielsen 5 3.15 Netherlands 76 Methodology 6 3.16 Norway 77 Authorship and Research Acknowledgements 7 3.17 Poland 16TH 78 3.18 Portugal 79

SECTION 1 3.19 Romania 80 Executive Summary and Key Findings by Nic Newman 9 3.20 Slovakia JUNE 81 3.21 Spain 82

SECTION 2 3.22 Sweden 83 Further Analysis and International Comparison 33 3.23 TIME)Switzerland 84 2.1 How and Why People are Paying for Online News 34 3.24 Turkey 85 2.2 The Resurgence and Importance of Email Newsletters 38 AMERICAS 2.3 How Do People Want the Media to Cover Politics? 42 3.25 United States 88 2.4 Global Turmoil in the Neighbourhood: 3.26 Argentina 89 Problems Mount for Regional and Local News (BRITISH47 3.27 Brazil 90 2.5 How People Access News about Climate Change 52 3.28 Canada 91 3.29 Chile 92 SECTION 3 00.01 3.30 Mexico 93 Country and Market Data 59 ASIA PACIFIC EUROPE 3.31 Australia 96 3.01 United Kingdom 62 3.32 Hong Kong 97 3.02 Austria 63 3.33 Japan 98 3.03 Belgium 64 3.34 Malaysia 99 3.04 Bulgaria 65 3.35 Philippines 100 3.05 CroatiaEMBARGOED: 66 3.36 Singapore 101 3.06 Czech Republic 67 3.37 South Korea 102 3.07 Denmark 68 3.38 Taiwan 103 3.08 Finland 69 AFRICA 3.09 France 70 3.39 Kenya 106 3.10 Germany 71 3.40 South Africa 107 3.11 Greece 72 3.12 Hungary 73 SECTION 4 3.13 Ireland 74 References and Selected Publications 109 3.14 Italy 75 4 / 5

Foreword Professor Rasmus Kleis Nielsen Director, Reuters Institute for the Study of Journalism (RISJ)

2020 The coronavirus crisis is having a profound impact not just on Our main survey this year covered respondents in 40 markets, our health and our communities, but also on the news media. including 24 in Europe, and four in Latin America (Brazil, Argentina, Most of the research presented in this year’s Digital News Report Chile, and Mexico) as well as the United States and Canada. In was collected before the virus hit in many countries, but findings Asia we have added the Philippines this year to our existing seven from the work we have carried out in the months since underline markets (Japan, South Korea, Taiwan,16TH Hong Kong, Malaysia, that the crisis is very likely to accelerate long-term structural Singapore, and Australia) and we are delighted to include Kenya for changes towards a more digital, more mobile, and more the first time, following the inclusion of South Africa in 2019. In the platform-dominated media environment. years to come we plan to continue to make this report even more comprehensive and moreJUNE international by covering more countries The bulk of this report is based on data collected by a survey of in the Global South. Having said that, as we use online polling more than 80,000 people in 40 markets and reflects media usage and need to make meaningful comparisons, we continue to focus in January/February just before the coronavirus hit many of these on countries with high internet penetration and which are either countries. But the key trends that we document here, including broadly democratic or generally compare themselves to countries changes in how people access news, low trust, and rising concern with a democraticTIME) tradition. about misinformation have been a backdrop against which journalists, editors, politicians, and public health officials have been battling to A report of this scale and scope is only possible due to collaboration reach ordinary people with key messages over the last few months. from our partners and sponsors around the world. We are proud to have the opportunity to work with a number of leading academics We know that this crisis has substantially increased the amount and top universities in the report, as well as media experts from and frequency of news consumption as well as influenced attitudes the news industry itself. Our partners have helped in a variety to the news media, at least temporarily. We’ve captured this in a (BRITISHof different ways, from checking questionnaires to helping to second set of polling data collected in April when the crisis was at analyse and interpret the results. Many are also organising events its peak in some countries. This has helped us to see the impact of or publishing country reports looking in more detail at specific the crisis in terms of sources of news and also reminded us of the issues facing their national media – adding wider value to this critical role that the news media play at times of national crisis, international project. including documenting that people who rely on00.01 news media are better informed about the virus than those who do not. Given the richness of the research, this report can only convey a small part of the data collected and work done. More detail While many media companies have been enjoying record audience is available on our website, www.digitalnewsreport.org, which figures, news fatigue is also setting in, and the short-term and contains slidepacks and charts, along with a licence that long-term economic impact of the crisis is likely to be profound – encourages reuse, subject to attribution to the Reuters Institute. advertising budgets are slashed and a recession looms, threatening news media, some of whom are struggling with adapting to a Making all this possible, we are hugely grateful to our sponsors: changing world. Against this background, this year’s report also Google News Initiative, BBC News, Ofcom, the Broadcasting focuses on the shift towards paying for online news in many Authority of Ireland, the Dutch Media Authority (CvdM), the countries across the world, with detailed analysis of progress in three Media Industry Research Foundation of Finland, the Fritt countries (theEMBARGOED: UK, USA, and Norway). What lessons can we learn Ord Foundation in Norway, the Korea Press Foundation, and about what could persuade more people to pay directly for news? Edelman UK, as well as our academic sponsors at the Hans Bredow Institute, the University of Navarra, the University of This year, our report carries important data about the extent to Canberra, the Centre d’études sur les médias, Quebec, Canada, which people value and trust local news, perhaps the sector most and Roskilde University in Denmark. We are particularly grateful vulnerable to the economic shocks that will inevitably follow the to the Open Society Foundations, our newest sponsor, which has health crisis itself. And we also explore the way people access helped us to expand the report to cover more countries in the news about climate change as well as attitudes to media coverage Global South this year. for the first time. We are also grateful to YouGov, our polling company, who did Given that the coronavirus is continuing to impact the media everything possible to accommodate our increasingly complex landscape at great pace, we have significantly reduced the requirements and helped our research team analyse and commentary on this year’s market- and country-based pages, though contextualise the data. we publish the key data that we collected in January/February in full. Reuters Institute for the Study of Journalism / Digital News Report 2020 6

Methodology

This study has been commissioned by the Reuters Institute • It is also important to note that online surveys rely on recall, for the Study of Journalism to understand how news is being which is often imperfect or subject to biases. We have tried consumed in a range of countries. Research was conducted by to mitigate these risks through careful questionnaire design YouGov using an online questionnaire at the end of January/ and testing. On the other hand, surveys can be a good way of beginning of February 2020. capturing fragmented media consumption across platforms • Samples in each country were assembled using nationally (eg social media, messaging, apps, and websites), and tracking representative quotas for age, gender, region, and education. activities and changes over time. 2020 The data were also weighted to targets based on census/ • We conducted two additional surveys this year. The first was industry accepted data. a detailed study of paying for online news where we surveyed • As this survey deals with news consumption, we filtered out around 4,000 respondents in the16TH United States and the United anyone who said that they had not consumed any news in the Kingdom and around 2,000 respondents in Norway. Polling past month, in order to ensure that irrelevant responses didn’t was conducted by YouGov in February 2020 using a similar adversely affect data quality. This category averaged around 3%. methodology as for the main survey. In April we conducted an additional survey with the Misinformation, Science and • We should note that online samples will tend to under-represent Media project runJUNE by the Reuters Institute in collaboration the consumption habits of people who are not online (typically with the Oxford Internet Institute and support from the older, less affluent, and with limited formal education). In this sense Oxford Martin School to understand the impact of the novel it is better to think of results as representative of online populations coronavirus on media consumption in six countries (UK, USA, who use news at least once a month. In a country like Norway this Germany,TIME) Spain, Argentina, and South Korea). Samples sizes is almost everyone (98%) but in Mexico it is only two-thirds (66%) were approximately 2,000 in the UK and Germany, and 1,000 and in South Africa just over half (55%). Our sample in Kenya only elsewhere. We have indicated occasions where data come includes those aged 18-54 due to difficulties in reaching older from these additional surveys next to the appropriate chart. people online. These differences mean we need to be cautious Where we compare the results from this survey to the DNR, when comparing results between countries. We have marked we have removed those that use news less than once a month countries with lower internet penetration or less representative to make the data more directly comparable. For more details online samples with an asterisk (*) in the table at the end(BRITISH of on the methodology for this survey, see the standalone report this section and have been careful in the report not to directly (Nielsen et al. 2020). Open questions were used in our surveys compare these countries on issues where we know that the sample and some user comments have been drawn from these and are difference would make results invalid (eg paying for news). used in the text. • It is important to note that some of our survey-based results REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM00.01 / FURTHER ANALYSIS • A fuller description of the methodology, panel partners, and will not match industry data, which are often based on different a discussion of non-probability sampling techniques can Slidemethodologies, 1 such as web-tracking. The accuracy of these be found on our website along with the full questionnaire approaches can be high, but they are subject to different (digitalnewsreport.org). limitations, meaning that data can also be partial or incomplete.

Country/market Sample Internet Country/market Sample Internet Country/market Sample Internet size penetration size penetration size penetration Europe Netherlands 2014 96% Chile* 2005 77% UK 2011 95% Norway 2010 98% Mexico* 2023 66% Austria 2005 88% Poland 2008 78% Asia Pacific Belgium EMBARGOED:2010 94% Portugal 2012 78% Australia 2131 89% Bulgaria* 2006 67% Romania* 2017 74% Hong Kong 2023 89% Croatia 2009 92% Slovakia 2018 85% Japan 2006 93% Czech Republic 2006 88% Spain 2006 93% Malaysia* 2013 81% Denmark 2061 98% Sweden 2091 96% Philippines* 2019 72% Finland 2050 94% Switzerland 2012 94% Singapore 2014 88% France 2038 92% Turkey* 2017 83% South Korea 2304 96% Germany 2011 96% Americas Taiwan 1027 93% Greece* 2015 73% USA 2055 89% Africa Hungary 2011 89% Argentina* 2007 93% Kenya (18-54)* 2003 87% Ireland 2006 92% Brazil* 2058 71% South Africa* 2006 55% Italy 2015 93% Canada 2002 93%

Source: Internet World Stats (http://www.internetworldstats.com). *Data are from more urban areas, rather than a fully nationally representative sample. These will tend to represent richer and more connected users which should be taken into consideration when interpreting results. 6 / 7

Authorship and research acknowledgements

Nic Newman is Senior Research Associate at the Reuters Institute for the Study of Journalism and is also a consultant on digital media, working actively with news companies on product, audience, and business strategies for digital transition. He also writes an annual report for the Institute on future media and technology trends. 2020

Dr Richard Fletcher is a Senior Research Fellow at the Reuters Institute for the Study of Journalism. He is primarily interested in global trends in digital news consumption, the use of social media by journalists and 16TH news organisations, and more broadly, the relationship between computer-based technologies and journalism. JUNE Dr Anne Schulz is a postdoctoral Research Fellow at the Reuters Institute for the Study of Journalism. Her doctoral work focused on populism, media perceptions, and news consumption. She is researching questions surrounding news audiences andTIME) digital news.

Dr Simge Andı is a postdoctoral Research Fellow at the Reuters Institute for the Study of Journalism. Her doctoral work explored online misinformation, mainly focusing on the sharing of false information via social media. She uses survey and(BRITISH experimental data to study the consumption and sharing of news.

Professor Rasmus Kleis00.01 Nielsen is Director of the Reuters Institute for the Study of Journalism, Professor of Political Communication at the University of Oxford, and served as Editor-in-Chief of the International Journal of Press/Politics from 2015 to 2018. His work focuses on changes in the news media, political communication, and the role of digital technologies in both.

Market-level commentary and additional insight around media developments have been provided by academic partners and by our EMBARGOED:network of Reuters Journalist Fellows around the world.1 Additional expert analysis and interpretation of the survey data were provided by the team at YouGov, in particular, Charlotte Clifford, Lucie Larboulette, David Eastbury, Mark Pellatt, and Anna Wilson.

1 Reuters Fellowships offer an opportunity to mid-career journalists to spend time researching an aspect of journalism for one or more terms at the Institute in Oxford. Reuters Institute for the Study of Journalism / Digital News Report 2020 8

2020

16TH

JUNE

TIME)

(BRITISH

00.01

EMBARGOED: 8 / 9

SECTION 1 Executive Summary and Key Findings

Nic Newman Senior Research Associate, 2020 Reuters Institute for the Study of Journalism 16TH

JUNE

TIME)

(BRITISH

00.01

EMBARGOED: Reuters Institute for the Study of Journalism / Digital News Report 2020 10

This year’s report comes in the midst of a global A SUMMARY OF SOME OF THE MOST IMPORTANT health pandemic that is unprecedented in modern FINDINGS FROM OUR 2020 RESEARCH times and whose economic, political, and social • The coronavirus crisis has substantially increased news consequences are still unfolding. The seriousness consumption for mainstream media in all six countries where of this crisis has reinforced the need for reliable, we conducted surveys before and after the pandemic had taken effect. Television news and online sources have seen accurate journalism that can inform and educate significant upticks, and more people identify television as populations, but it has also reminded us how their main source of news, providing temporary respite from a open we have become to conspiracies and picture of steady decline. Consumption of printed newspapers has fallen as lockdowns undermine physical distribution, misinformation. Journalists no longer control almost certainly accelerating the shift to an all-digital future. access to information, while greater reliance • At the same time, the use of online and social media on social media and other platforms give people substantially increased in most countries. WhatsApp saw access to a wider range of sources and ‘alternative the biggest growth in general with increases2020 of around ten facts’, some of which are at odds with official percentage points in some countries, while more than half of those surveyed (51%) used some kind of open or closed advice, misleading, or simply false. online group to connect, share information, or take part in a local support network. Much of the data in this publication was collected before the virus 16TH hit many of the countries featured in this survey, so to a large • As of April 2020, trust in the media’s coverage of COVID-19 extent this represents a snapshot of these historic trends. But was relatively high in all countries, at a similar level to to get a sense of what has changed, we repeated key parts of our national governments and significantly higher than for survey in six countries (UK, USA, Germany, Spain, South Korea, individual politicians.JUNE Media trust was more than twice the and Argentina) in early April. These responses confirm industry level for social networks, video platforms, or messaging data which show increased consumption of traditional sources of services when it came to information about COVID-19. news, especially television, but also some online news sources. From ourTIME) wider dataset collected in January: Journalism matters and is in demand again. But one problem • Global concerns about misinformation remain high. Even before for publishers is that this extra interest is producing even less the coronavirus crisis hit, more than half of our global sample income – as advertisers brace for an inevitable recession and said they were concerned about what is true or false on the print revenue dips. Against this background it is likely we’ll see internet when it comes to news. Domestic politicians are the a further drive towards digital subscription and other reader single most frequently named source of misinformation, though payment models which have shown considerable promise in the in some countries – including the United States – people who last few years. This is what makes it so important to understand(BRITISH self-identify as right-wing are more likely to blame the media – how various models are progressing in different markets, such part of a ‘pick-your-side’ dynamic. Facebook is seen as the main as the United States, Norway, and the UK, where we have channel for spreading false information almost everywhere but conducted more in-depth research on pay this year. While the WhatsApp is seen as more responsible in parts of the Global coronavirus crisis is likely to dramatically affect00.01 the short-term South such as Brazil and Malaysia. prospects for many publishers, our findings provide long-term insights into one important part of the future of the business • In our January poll across countries, less than four in ten of news. We also ask about the implications for society if more (38%) said they trust most news most of the time – a fall of high-quality information disappears behind paywalls, a dilemma four percentage points from 2019. Less than half (46%) said that has become more real during this health emergency. they trust the news they use themselves. Political polarisation linked to rising uncertainty seems to have undermined trust Looking to the future, publishers are increasingly recognising in public broadcasters in particular, which are losing support that long-term survival is likely to involve stronger and deeper from political partisans from both the right and the left. connection with audiences online, which is why we have also • Despite this, our survey shows that the majority (60%) still examined the growing importance of emails and podcasts, prefer news that has no particular point of view and that only formats that are being deployed in greater numbers to increase EMBARGOED: a minority (28%) prefer news that shares or reinforces their engagement and loyalty. The coronavirus crisis has presented an views. Partisan preferences have slightly increased in the immediate and urgent existential threat, but in the section on United States since we last asked this question in 2013 but climate change, we explore how audiences access news about even here a silent majority seems to be looking for news that this longer-term threat to human existence and how they feel at least tries to be objective. about media coverage. • As the news media adapt to changing styles of political Our report this year, based on data from six continents and communication, most people (52%) would prefer them to 40 markets, aims to cast light on the key issues that face the prominently report false statements from politicians rather than industry at a time of unprecedented uncertainty. The overall not emphasise them (29%). People are less comfortable with story is captured in this Executive Summary, followed by political adverts via search engines and social media than they Section 2 with chapters containing additional analysis, and then are with political adverts on TV, and most people (58%) would individual country and market pages in Section 3 with detailed prefer platforms to block adverts that could contain inaccurate data and extra context. claims – even if it means they ultimately get to decide what is true. 10 / 11

• We have seen significant increases in payment for online news CORONAVIRUS REMINDS PEOPLE OF THE VALUE in a number of countries including the United States 20% OF TRADITIONAL NEWS SOURCES (+4) and Norway 42% (+8), with smaller rises in a range of other markets. It is important to note that across all countries Over the last nine years, our data have shown online news most people are still not paying for online news, even if some overtaking television as the most frequently used source of news publishers have since reported a ‘coronavirus bump’. in many of the countries covered by our online survey. At the same time, printed newspapers have continued to decline while • Overall, the most important factor for those who subscribe social media have levelled off after a sharp rise. is the distinctiveness and quality of the content. Subscribers believe they are getting better information. However, a large The coronavirus crisis has significantly, though almost certainly number of people are perfectly content with the news they temporarily, changed that picture. Television news has seen an can access for free and we observe a very high proportion of REUTERSuplift in INSTITUTEall six countries FOR THE whereSTUDY OFwe JOURNALISMpolled in both / FURTHER January ANALYSISand April non-subscribers (40% in the USA and 50% in the UK) who 2020. Taking Germany as an example (see chart), a 12-point decline say that nothing could persuade them to pay. Slide 2 REUTERSin reach INSTITUTEfor TV news FOR was THE partially STUDY OF reversed JOURNALISM as many / FURTHER people ANALYSIS turned • In countries with higher levels of payment (eg the USA and Slideto trusted sources3 of news including public service media. Norway) between a third and half of all subscriptions go to just 2020 a few big national brands – suggesting that winner-takes-most PROPORTION THAT USED EACH AS A SOURCE OF NEWS dynamics are persisting. But in both these countries a significant IN THE LAST WEEK (2013–20) – GERMANY minority are now taking out more than one subscription, TV Print (incl. mags) Online (incl. social media) Social media often adding a local or specialist publication. 16TH Jan Apr • In most countries, local newspapers and their websites 2020 2020 100% remain the top source of news about a particular town Social media or region, reaching four in ten (44%) weekly. But we find 82% JUNE Online (incl. social media) that Facebook and other social media groups are now 75% 70% 72% 63% 69% used on average by around a third (31%) for local news and 63% 70% Print information, putting further pressure on companies and 50% TV their business models. TIME) 37% 39% • Access to news continues to become more distributed. 25% 33% 26% 18% Across all countries, just over a quarter (28%) prefer to start

their news journeys with a website or app. Those aged 18–24 0% (so-called Generation Z) have an even weaker connection with 2013 2014 2015 20172016 2018 2019 Jan Apr 2020 2020 websites and apps and are more than twice as likely to prefer to access news via social media. Across age groups, use of Q3. Which, if any, of the following have you used in the last week as a source of news? Q4. (Apr. 2020). Which, if any, of the following have you used in the last week as a source of news? Total Instagram for news has doubled since 2018 and looks likely(BRITISH 2013–20 samples ≈ 2000. Note. Apr. 2020 figures adjusted to exclude non-news users for comparability. to overtake Twitter over the next year. • To counter the move to various platforms, publishers have been looking to build direct connections with consumers Weekly TV news consumption rose by an average of five percentage via email and mobile alerts. In the United00.01 States one in five points across all six countries. But it is worth noting that social (21%) access a news email weekly, and for almost half of media were also substantially up (+5) as more people used these these it is their primary way of accessing news. Northern networks for finding and sharing news in combination with European countries have been much slower to adopt email television and online sites. By contrast, the lockdowns hit the reach news channels, with only 10% using email news in Finland. of print newspapers and magazines with a six-point drop in Spain, not helped by difficulties in distributing physical copies. • The proportion using podcasts has grown significantly in the last year, though coronavirus lockdowns may have temporarily reversed this trend. Across countries, half of all PROPORTION THAT USED EACH AS A SOURCE OF NEWS respondents (50%) say that podcasts provide more depth and IN THE LAST WEEK (APRIL 2020) – SELECTED COUNTRIES understanding than other types of media. Meanwhile, Spotify has becomeEMBARGOED: the number one destination for podcasts in a Online TV Social Radio Print (incl. social media (incl. mags) number of markets, overtaking Apple’s podcast app. media) UK 79% 71% 47% 35% 18% • Overall, almost seven in ten (69%) think climate change is a serious problem, but in the United States, Sweden, and USA 73% 60% 47% 21% 16% Australia a significant minority dispute this. This group Germany 69% 72% 39% 41% 26% tends to be right-wing and older. Younger groups access Spain 83% 71% 63% 24% 28% much of their climate change news from social media and South Korea 85% 65% 51% 14% 19% by following activists like Greta Thunberg. Argentina 90% 77% 78% 24% 30% Average change +2 +5 +5 +2 -2 from January

Q4. (Apr. 2020). Which, if any, of the following have you used in the last week as a source of news? Base: Total sample: UK = 2191, USA = 1221, Germany = 2003, Spain = 1018, South Korea = 1009, Argentina = 1003. Note. Figures adjusted to exclude non-news users for comparability. 91.7001 mm

100% Social media Reuters Institute for the Study of Journalism / Digital News Report 2020 REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS 12 Online (incl. social media) 75% Slide Print5

50% TV

The25% change of underlying preferences is even more clear when we PROPORTION THAT USED EACH AS A SOURCE OF NEWS ask people to choose their main source of news. The UK shows IN THE LAST WEEK (APRIL 2020) – UK a 20-percentage-point switch in preference from online to TV 0% Jan Apr 100% between the end2013 of January2014 2015 and the start20172016 of2018 April,2019 and2020 JanArgentinaApr2020 +7 2020 2020 19 points, with the average across six countries being a change 90 +7 of 12 points. +1 +12 80% -1 81 75 76 In the UK, Prime Minister Boris Johnson’s address telling Britons +9 68 +25 to stay at home was one of the most-watched broadcasts in UK 60% 61 television history, with 27m tuning in live (excluding live streams 57 REUTERSvia news INSTITUTE websites FOR and THE apps). STUDY2 Nightly OF JOURNALISM viewership / FURTHER of BBC ANALYSIS TV +8 +6 Slidebulletins was 4 up by around 30% in March while the European 40% 42 -7 Broadcasting Union (EBU) reported daily viewing increases 35 -5 across countries of 14% in the early stages.3 30 20% REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS +1 21 2020 Slide 6 12 0% Online Social media Print TV (inc. social media) 16THU35 35+ 55+ UK TRAFFIC DOUBLES TO BBC NEWS WEBSITE JAN- APRIL 2020 Q4. (Apr. 2020). Which of the following have you used in the last week as a source of news? Base: U35/35+/55+: UK = 548/1643/923. Note. Showing change from Jan. 2020. Figures adjusted SOCIAL DISTANCING AND LOCKDOWN 35,000,000 to exclude non-news users for comparability. 30,000,000 Number of daily JUNE 25,000,000 browsers initially doublesBy contrast to the UK experience, in countries with fewer before falling back 20,000,000 widely used and broadly trusted sources, the television 15,000,000 experience has been very different, with CNN and Fox 12 March 6 April 10,000,000 NewsGovernment covering announces President Trump’sBoris Johnson press admitted conferences in Šrst shielding measures to hospital 5,000,000 fundamentallyTIME) different ways, and some broadcasters Two children watch British Prime Minister Boris Johnson’s press conference as the spread of coronavirus0 disease (COVID-19) continues in Hertford, Britain March 23, 2020. REUTERS/ refusing to transmit live from sessions where the President Andrew Couldridge01 Jan 08 Jan 15 Jan 22 Jan 29 Jan 05 Feb 12 Feb 19 Feb 26 Feb 04 Mar 11 Mar 18 Mar 25 Mar 01 Apr 08 Apr 15 Apr 22 Apr 29 Apr 2020 2020 2020 2020 2020 2020 2020 2020 2020 often2020 goes2020 directly2020 against2020 medical2020 2020and scientific2020 2020 advice.2020 In the United States we see less of a boost for television news SOURCE: BBC. Note: Unique browsers does not necessarily equate to people as they can be related to differentoverall, devices or thoughweb browsers a similar growth in usage from under-35s. The uplift in TV and social media was experienced across all age groups, with under-35s proportionally showing the biggest(BRITISH Industry data also indicate strong traffic increases for increase in use of television as well as for using social media to online news with the most trusted brands often benefitting access news. In the UK, under-35s showed an increase in watching disproportionately. The BBC reported its biggest week ever TV news of 25 percentage points compared with January, though for UK visitors, with more than 70m unique browsers as the their core preference for online and social media remained. Older lockdown came into effect – though traffic has since returned groups read fewer newspapers and watched more00.01 television news. to more normal levels.

UK TRAFFIC DOUBLES TO BBC NEWS WEBSITE JAN-APRIL 2020

SOCIAL DISTANCING AND LOCKDOWN 35,000,000

30,000,000 Number of daily 25,000,000 browsers initially doubles EMBARGOED: before falling back 20,000,000

15,000,000 12 March 6 April 10,000,000 Government Boris Johnson announces
rst admitted shielding measures to hospital

Number of daily UK unique browsers 5,000,000

Two children watch0 British Prime Minister Boris Johnson’s press conference as the spread of coronavirus disease (COVID-19) continues in Hertford, Britain March 23, 2020. REUTERS/Andrew Couldridge 01 Jan 08 Jan 15 Jan 22 Jan 29 Jan 05 Feb 12 Feb 19 Feb 26 Feb 04 Mar 11 Mar 18 Mar 25 Mar 01 Apr 08 Apr 15 Apr 22 Apr 29 Apr 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020 2020

SOURCE: BBC. Note: Unique browsers does not necessarily equate to people as they can be related to different devices or web browsers.

2 https://www.theguardian.com/tv-and-radio/2020/mar/24/boris-johnsons-covid-19-address-is-one-of-most-watched-tv-programmes-ever 3 https://www.newstatesman.com/culture/tv-radio/2020/03/coronavirus-bbc-corporation-boris-johnson-protect. EBU Report, Covid-19: The Impact on Digital Media Consumption, Apr. 2020, https://www.ebu.ch/publications 91.7001 mm

REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER 12ANALYSIS/ 13 Slide 5 Slide 9

PROPORTION THAT USED EACH AS A SOURCE OF NEWS Commercial media have also reported a significant increase in TRUST IN DIFFERENT SOURCES OF NEWS AND INFORMATION IN THE LAST WEEK (APRIL 2020) – UK REUTERSonline traffic INSTITUTE and FOR have THE produced STUDY OF JOURNALISM many examples / FURTHER of innovations ANALYSIS ABOUT CORONAVIRUS (APRIL 2020) – AVERAGE OF SIX in data journalism and other visual formats online to help explain COUNTRIES 100% +7 Slide 7 the crisis. Websites and apps have included practical guides for Trust Neither/Don’t know Don’t trust 90 +7 managing isolation and staying safe. +1 +12 Scientists, doctors 83% 5% 80% -1 81 75 76 EXPLANATORY AND DATA-RICH ONLINE COVERAGE FROM +9 National health organisations 76% 9% 68 +25 THE WASHINGTON POST 60% 61 57 Global health organisations 73% 12% +8 +6 News organisations 59% 18% 40% 42 -7 35 -5 30 The national government 59% 23% 20% +1 21 Ordinary people I know personally 43% 21% 12 2020 0% Individual politicians 35% 35% Online Social media Print TV Overall our April 2020 survey found the news media were (inc. social media) considered to have done a good job in helping ordinary people Ordinary people that I do not U35 35+ 55+ 17% 45% understand the extent of the crisis (60%), and also in making know personally 16TH clear what people can do personally to mitigate the impact Q4. (Apr. 2020). Which of the following have you used in the last week as a source of news? Search engines Base: U35/35+/55+: UK = 548/1643/923. Note. Showing change from Jan. 2020. Figures adjusted (65%). Though some media have in the past been accused of 45% 18% to exclude non-news users for comparability. (Google, Bing, Naver) sensationalising stories, on this occasion only a third (32%) REUTERSthink that INSTITUTE the media FOR THEhave STUDY exaggerated OF JOURNALISM the severity / FURTHER of ANALYSIS the Video sites (YouTube, Vimeo)JUNE 29% 34% situation, though concern was higher in the United States Slide 8 Social media (38%) and Argentina (41%), and amongst those that distrust eg Facebook, Twitter) 26% 40% the media already. Messaging apps (eg WhatsApp, 24% 40% TIME)Facebook Messenger) ATTITUDES TOWARDS NEWS MEDIA COVERAGE OF CORONAVIRUS (APRIL 2020) – SELECTED COUNTRIES Q10/11. (Apr. 2020). How trustworthy would you say news and information about coronavirus (COVID-19) from the following is? Please use the scale below, where 0 is ‘not at all trustworthy’ and 10 is ‘completely trustworthy’. Base: Total sample: USA, UK, Germany, Spain, Argentina, South Korea = 8522. Note. Trust = 6–10, Don’t trust = 0–4.

At around the peak of the lockdowns, trust in news organisations (BRITISHaround COVID-19 was running at more than twice that for social media, video sites, and messaging applications where around four 60% 65% 32% in ten see information as untrustworthy. By contrast information Agree media Agree media has Agree media has has helped me helped explain exaggerated found in search engines was seen as more reliable. Of course, neither understand the crisis what I can do 00.01the impact search nor social companies create content themselves, so trust in this context is a reflection of the selection decisions they make.

Least Least Most It is particularly striking that average levels of trust in the 52% in 57% in 41% in national government and news organisations are almost USA South Korea Argentina identical, perhaps reflecting the way that in the early stages of this crisis many media organisations focused on amplifying government messages about health and social distancing, including carrying extended government press conferences. As Q14. (Apr. 2020). To what extent do you agree or disagree with the following statements about the things return to normal, the media are likely to become more coronavirus (COVID-19)? Base: Total sample: USA, UK, Germany, Spain, Argentina, South Korea = 8522. critical of government and this may in turn lead to a return of EMBARGOED: more partisan approaches to media trust.

And in terms of trust for information about coronavirus, national news organisations score relatively well, behind doctors and health organisations but ahead of individual politicians and ordinary people. In recent years some populist politicians in particular have taken to undermining the media but this coronavirus pandemic has been a reminder that even weakened media play a critical role in informing populations and shaping opinion. REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS Slide 10 Reuters Institute for the Study of Journalism / Digital News Report 2020 REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS 14 Slide 11

SOCIAL MEDIA AND THE CRISIS PROPORTION OF 18-24s THAT USED EACH AS A SOURCE OF CORONAVIRUS NEWS IN THE LAST WEEK (APRIL 2020) – While people are wary of social media and messaging apps, SELECTED COUNTRIES high usage may suggest that many people are confident in their own ability to spot misinformation or just that they value these platforms for other reasons, such as keeping in touch with family or friends, and for mutual support. Instagram Snapchat TikTok

In April we found that across our six surveyed countries almost a UK 24% 19% 6% quarter (24%) used WhatsApp to find, discuss, or share news about COVID-19 – up seven points on average on our January survey which USA 26% 14% 11% asked about usage for any news. Around a fifth (18%) joined a People applaud from their balcony as part of an event organised via social media Germany 38% 10% 8% supportto thank Spanishor discussion healthcare group workers. with Malaga, people Spain. REUTERS/Jon they didn’t Nazca know on either Facebook or WhatsApp specifically to talk about COVID-19 and Spain 26% 1% 1% half (51%) took part in groups with colleagues, friends, or family. South Korea 10% 1% 5% One in ten accessed closed video chats using platforms like Zoom, 2020 Houseparty, and Google Hangouts – many for the first time. Argentina 49% 7% 9%

Q8. (Apr. 2020). Which, if any, of the following have you used in the last week as a source of news or information about coronavirus (COVID-19)? Base: 18-24s: UK = 202, USA = 153, Germany = 183, Spain = 80, South Korea = 83, Argentina = 151. 16TH

LASTING EFFECTS?

Media habits changedJUNE significantly during the COVID-19 lockdowns even if levels of interest have proved hard to sustain.4 More people turned to live broadcast television news and to trusted news sources online, but the lockdowns have also acceleratedTIME) the use of new digital tools, with many people joining online groups or taking part in video conferencing for the first time. Together with the impact on print production and distribution, it is likely that the net effect will be to speed up rather than slow down the shift to digital.

The implications for trust are harder to predict. In most countries (BRITISHthe media played a largely supportive role in the early stages of the crisis – when lives were most at risk. But that consensus

People applaud from their windows and balconies as they take part in an event organized has already started to break down as normal activities resume through social media to show gratitude to healthcare workers during a partial lockdown as and disagreements resurface about the best way to manage the part of a 15-day state of emergency to combat the spread of coronavirus disease (COVID-19), outside a pharmacy in Ronda, southern Spain, March 20, 2020. REUTERS/Jon00.01 Nazca recovery. Any ‘trust halo’ for the media may also be short-lived.

Meanwhile Instagram and Snapchat have become popular with younger groups for accessing news about COVID-19. Celebrities and influencers play an outsized role on these networks, with some sharing music, running exercise classes as well as commenting on the wider health issues. Almost half of our 18–24 respondents in Argentina (49%) used Instagram, 38% in Germany. One in ten (11%) accessed COVID-19 news via TikTok in the US and 9% in Argentina.

Social media may be helping to spread false and misleading information, butEMBARGOED: it has also supported people at a time of anxiety and isolation and provided an effective way to amplify reliable information. It is important to remember that social media are generally used in combination with other types of information. In the six countries we surveyed in April, more than four-in-ten (43%) accessed both social media for news and traditional media sources on a weekly basis. A more detailed analysis showed that, in most countries, those who relied on news organisations for information about coronavirus were more knowledgeable about the crisis, whereas there was no consistent, significant pattern for those who relied on various platforms including social media – they were not more informed but, despite some fears to the contrary, they were not misinformed either (Nielsen et al. 2020).

4 https://www.niemanlab.org/2020/04/the-coronavirus-traffic-bump-to-news-sites-is-pretty-much-over-already 14 / 15

TRUST IN THE NEWS MEDIA CONTINUES TO FALL Divided societies seem to trust the media less, not necessarily GLOBALLY because the journalism is worse but because people are generally dissatisfied with institutions in their countries and perhaps As the coronavirus hit, we observed overall levels of trust in the 75%because news outlets carry more views that people disagree with. Brexit Divisive 2019 Right news at their lowest point since we started to track these data. vote election REUTERSIn a direct INSTITUTE comparison FOR THE with STUDY 2019 OF we JOURNALISM find that / FURTHERfewer than ANALYSIS four in One recent58% example of how this works came in the UK electionLeft in December 2019 where Boris Johnson asked the people to Slideten (38%) say12 they trust most news most of the time – down four 50% percentage points. Less than half (46%) say they trust the news endorse46% his Brexit deal. The Tory victory came after a toxic political that they themselves use. campaign where the media were heavily36% criticised by both sides (Fletcher et al. 2020). Many on the left blamed the Labour defeat 25% on unfair media treatment of its leader Jeremy Corbyn by a ‘biased PROPORTION THAT TRUST EACH MOST OF THE TIME – press’. The impact can be seen in the next15% chart with trust amongst ALL MARKETS respondents who self-identify on the left collapsing from 38% in 0% Trust news Trust news Trust news Trust news January 2019 to just 15% a year later. Trust202020192018201720162015 amongst those on the overall I use in search in social right also declined but by much less. 2020 REUTERSIn the United INSTITUTE States FOR we THE observe STUDY the OF JOURNALISM same picture / FURTHER but in reverse. ANALYSIS 38% 46% 32% 22% SlideIt is the right 16 that has lost trust, partly inflamed by the anti-media (-4) (-3) (-1) (-1) rhetoric of the president himself. A bruising Democratic Presidential Primary may help account for the 16THdecline in trust on the left.

Mostly this is about trust in mainstream Uncertainty in distributed environments, PROPORTION THAT TRUST MOST NEWS MOST OF THE TIME media and in the sources that people use information unchecked, hard to BY LEFT-RIGHT POLITICAL LEANING (2015–20) – UK & USA distinguish news from rumour... 75% JUNE Brexit Divisive 2019 Right Q6_1/2/3/4. Please indicate your level of agreement with the following statements: I think vote election UK you can trust most news/the news I use/news in search/news in social media most of the time. 58% Le Base: Total sample: All markets = 80155. 50% 46% Trust in the news overall 36% We continue to see considerable country differences, ranging 25% TIME) 28% 75% from Finland and Portugal where over half (56%) say they trust 15% (-12) most news most of the time, to less than a quarter in Taiwan 0% (24%), France (23%), and South Korea (21%). Just six countries 202020192018201720162015 Left Right 59 58 57 50%now have trust levels53 of more than 50%. 52 50 49 49 48 48 48 47 46 46 46 46 45 44 75% 43 42 Trump New election Right Notable changes over the last 12 months include a 16-percentage 40 40 40 40 39 39 39 39 (BRITISH elected 35cycle USA point fall in Hong Kong (30%) following violent street protests 33 33 32 Le 25% 31 over a proposed extradition law. In Chile, which has seen regular 50% 28 28 27 Trust in the24 22 39% news overall demonstrations about inequality, the media has lost trust (-15) 35% partly because it is seen as too close to the elites and not focusing 25% 25% 29% 0% 5 its attention enoughNLDDENPORFIN on underlyingPOLZAFBELMEXCAN grievances. TURHKSUIGERIREBRA NOR CHL AUS SPA BGRUKSGP ITA CRO JPN ARG AUT SWE ROU CZE SVK USA MYS TWN HUN FRAGRE KOR 00.01 13% (-3)

There were also significant falls in the United Kingdom (-12), 0% 202020192018201720162015 Left Right Mexico (-11), Denmark (-11), Bulgaria (-7), Canada (-8), and REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS Australia (-6) where our poll coincided with bitter debates over Q6_2016_1. Please indicate your level of agreement with the following statement: I think you can the handling of some of Australia’s worst-ever bush fires. trust most news most of the time. Q1F. Some people talk about ‘left’, ‘right’ and ‘centre’ to describe Slide 13 parties and politicians. With this in mind, where would you place yourself on the following scale? Base: 2015–20 Left/Right: USA ≈ 500/500, UK ≈ 500/300.

PROPORTION THAT AGREE THEY CAN TRUST MOST NEWS MOST OF THE TIME – ALL MARKETS EMBARGOED: 75% -3 Denmark Mexico Chile Hong Kong UK Concern about economy Violent Violent Divisive election, and gang violence street street Brexit, attacks -11 -8 protests protests on media 56 56 55 50% -11 52 51 50 -6 48 48 46 -7 45 45 45 45 44 44 -15 -16 -11 -3 -12 40 39 39 38 38 38 37 36 36 -4 33 33 33 30 30 25% 29 29 28 28 28 27 27 25 24 23 21

0% AUT SUICANBELPOLNOR GERDENIREZAKENBRA NLDTURPORFIN GERDENIREZAKENBRA SUICANBELPOLNOR AUT CRO MEX SWE AUS ROU JPN SPA CZESGN ARG HKCHLBGR ITA USA GRE SVK UK PHLHUN MYS FRATWN KOR

Q6_1. Please indicate your level of agreement with the following statement: I think you can trust most news most of the time. Base: Total sample in each market ≈ 2000, Taiwan = 1027.

5 https://www.aljazeera.com/programmes/listeningpost/2019/11/chile-protests-media-191103105957626.html Reuters Institute for the Study of Journalism / Digital News Report 2020 16

Although the left has long had a problem with right-wing bias in the Looking across nine markets we see that the majority in each UK press, the Brexit crisis saw the public service BBC increasingly country say they prefer news with no particular point of view. In a in the firing line. The BBC has a duty to be duly impartial in its news sense this is not surprising given that traditional expectations are coverage but our data show that, although overall trust in its news that journalists should produce neutral and detached news, but REUTERSremains INSTITUTEhigh, criticism FOR THE from STUDY those OF who JOURNALISM are politically / FURTHER committed ANALYSIS the differences between countries are striking. Slide(on the left 17 and the right) has grown over the last few years. Trust in the BBC with the most partisan groups has fallen by 20 percentage This preference for neutral news is strongest in Germany, Japan, points since 2018. the UK, and Denmark – all countries with strong and independent publicREUTERS broadcasters. INSTITUTE FOR A THE preference STUDY OF JOURNALISM for more partial / FURTHER news ANALYSIS is strongest in Spain, France, and Italy – countries that scholars PROPORTION THAT TRUST BBC NEWS BY LEVEL Slidehave labelled 18 ‘polarised pluralist’ (Hallin and Mancini 2004) – OF POLITICAL COMMITMENT (2018 AND 2020) – UK as well in the United States. 100% More Partisan -9 PROPORTION THAT SAY THEY PREFER TO GET NEWS -8 Partisan 75% 79 FROM SOURCES THAT HAVE NO POINT OF VIEW2020 – 71 -20 SELECTEDCentrist COUNTRIES 67 70 63 50% 15 47 Germany 16TH 80 Challenge your point of view 25% 5

17 0% Trust BBC 2018 Trust BBC 2020 Japan JUNE 78 Centrist Partisan More Partisan 6 No point of view 13 Q6_2018_trust. How trustworthy would you say news from the following brands is? Please use the scale below, where 0 is ‘not at all trustworthy’ and 10 is ‘completely trustworthy’. Q1F. UK 76 Some people talk about ‘left’, ‘right’, and ‘centre’ to describe parties and politicians. With this in mind, where would you place yourself on the following scale? Base: Centrist/Partisan/More TIME)11 partisan: 2018 = 1018/645/170, 2020 = 1019/536/97. Note centrist = those that selected ‘centre’, ‘slightly left/right of centre’, Partisan = ‘fairly left/right wing’, More partisan = ‘very left/right wing’. 14 Share your point of view Denmark 68 Public service media remain by and large the most trusted brands, 18 especially in Northern European countries where they have a strong 34 tradition of independence. But criticisms from the extremes do (BRITISHSpain 55 seem to be chipping away at this confidence in many countries, especially when combined with anti-elitist rhetoric from populist 11 politicians. Our country and market-based pages clearly show 22 that, though trust remains high, distrust of public service media is Italy 65 growing and is often higher than for many other00.01 news brands. 13 For full brand trust scores see section 3: Country and Market pages 30 USA 60 DESPITE PARTISAN PRESSURES, SILENT MAJORITY 10 STILL PREFER ‘OBJECTIVE’ NEWS 20 Greater political polarisation has coincided with an explosion France 58 of low-cost internet publishing which in turn has led to the 22 widespread availability of partisan opinions online. With news coverage increasinglyEMBARGOED: commoditised, parts of the traditional 43 media have also focused more on strong and distinctive Brazil 51 opinion as a way of attracting and retaining audiences. Some 6 commentators have increasingly questioned the value of objective news in a world where people have ready access to news 0% 20% 40% 60% 80% 100% from so many different points of view, while others worry that Share your point of view social media and algorithms are encouraging echo-chambers and No point of view pushing communities apart. In this context we were interested Challenge your point of view to know if consumer preferences for news that reinforce people’s views had grown since we last looked at this subject in 2013. Q5c_2020. Thinking about the different kinds of news available to you, do you prefer getting news from sources that share/challenge/have no point of view. Base: Total sample (excluding don’t knows): USA = 1760, UK = 1659, Germany = 1624, France = 1437, Italy = 1762, Spain = 1756, Denmark = 1611, Japan = 1538, Brazil = 1901. 16 / 17

Comparing 2020 with data from 2013, we see increased preference THE ROLE OF THE NEWS SOURCE 100% over time in the UK (+6) for news that has ‘no particular point of Challenge your point of view view’. At the same time, the proportion that prefers news that It is interesting how people’s attitudes to objective, partial, or challenging news is related to the source of news they use most ‘shares their point of view’ has declined by a similar amount (-6). No point of view It is hard to be sure about the reasons, but one possibility is that a often.75% In the next chart we can see that in the USA people who silent majority is reacting against a perceived increase in agenda- regularly use strident 24-hour news channels like Fox and CNN Share your point of view are most likely to prefer partial news, followed by social media filled, biased, or opinion-based news. These themes came out 61 strongly in comments from our survey respondents. and then print. 50% 48 I prefer that news is delivered objectively so individuals In the UK we see47 the reverse picture, with people who use are able to analyse what they have heard/read television40 as main source most likely to prefer neutral or objective news. This is not surprising given that television in and come to their own conclusions without being 25% 30 unduly influenced. the UK has an obligation to be duly impartial and26 is regulated Female, 61, UK accordingly. By contrast, those whose main source is a printed REUTERSnewspaper INSTITUTE (or social FOR media)THE STUDY are OF three JOURNALISM times as / FURTHER likely to ANALYSIS prefer 7 7 5 news0% that shares their point of view. In Spain and Brazil, we see SlideTV 20 news (24 hour) Social media 2020Newspapers A DIFFERENT STORY IN THE UNITED STATES all three media sources being associated with a preference for partial news – and social media playing the biggest role. In the United States, where both politics and the media have become increasingly partisan over the years, we do find an increase PROPORTION THAT SAY THEY PREFER16TH TO GET NEWS FROM in the proportion of people who say they prefer news that shares SOURCES THAT HAVE NO POINT OF VIEW BY MAIN SOURCE REUTERStheir point INSTITUTE of view FOR– up THE six STUDYpercentage OF JOURNALISM points since / FURTHER 2013 to ANALYSIS 30%. OF NEWS – SELECTED COUNTRIES ThisSlide is driven 19 by people on the far-left and the far-right who have both increased their preference for partial news sources. USA JUNE

100% Share your point of view Challenge your point of view No point of view PARTIAL TV AUDIENCES Challenge your point of view No point of view 75% TIME)

62% 34% 66 Share your point of view of those who self of those who self 56 identify on the identify on the 50% right watch Fox left watch CNN 51 (11% from the left) (9% from the right) 42 (BRITISH 35 25% 28 In the USA, audiences for cable television networks, such as Fox on the right and CNN and MSNBC on the left, are strongly weighted to one view or another. At the same time, the influence 7 8 6 0% of partisan websites (eg Breitbart, the Blaze, the00.01 Daily Caller, TV news (24 hour) Social media Printed newspapers Occupy Democrats, Being Liberal) grew rapidly ahead of the 2016 election and even today around a quarter (23%) of our sample UK visits at least one of these each week. In our survey we asked respondents who said they prefer news that ‘shares their point of 100% Share your point of view Challenge your point of view view’ to explain why they held that position. Many suggested that No point of view it helped avoid political arguments or because they felt they were Challenge your point of view 81 No point of view getting closer to the truth: 75%

I'm tired of hearing fake news from the dishonest 69 70 Share your point of view left socialistEMBARGOED: commie traitors!!! Female, 69, USA survey respondent 50%

Because CNN, MSNBC tell the truth whereas Fox News is not bound by pure facts. 25% Female, 73, USA survey respondent 20 21 Complaints like this about the media have become increasingly 11 11 8 9 vocal in recent years and yet it is striking that even in the United 0% States a majority (60%) still express a preference for news TV news Social media Printed newspapers without a particular point of view. We can also note that, during Q5c_2020. Thinking about the different kinds of news available to you, do you prefer getting the coronavirus crisis, partisan websites and TV brands showed news from sources that share/challenge/have no point of view. Base: Total sample (excluding stagnant or low traffic growth compared with other brands.6 don’t knows): USA = 1760, UK = 1659.

6 New York Times analysis of traffic changes https://www.nytimes.com/interactive/2020/04/07/technology/coronavirus-internet-use.html REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS Slide 22

PROPORTION THAT SAY THEY PREFER TO GET NEWS FROM SOURCES THAT HAVE NO POINT OF VIEW BY AGE – ALL MARKETS

32 18-24 25-34 34 35-44 Share point 31 45-54 of view 26 55+ 24 51 52 No point 57 of view 62 66 17 14 Challenge 12 point of view 11 10

0% 25% 50% 75%

Q5c_2020. Thinking about the different kinds of news available to you, do you prefer getting news from sources that share/challenge/have no point of view. Base: 18–24 = 8639, 25–34= 12603, 35–44 = 12643, 45–54 = 11605, 55+ = 23773. ReutersREUTERS Institute INSTITUTE for the FOR Study THE of Journalism STUDY OF JOURNALISM/ Digital News / FURTHER Report 2020 ANALYSIS 18 Slide 21

PROPORTION THAT SAY THEY PREFER TO GET NEWS FROM PROPORTION THAT SAY THEY PREFER TO GET NEWS SOURCES THAT HAVE NO POINT OF VIEW BY MAIN SOURCE FROM SOURCES THAT HAVE NO POINT OF VIEW BY AGE – OF NEWS – SELECTED COUNTRIES ALL MARKETS

18-24 32 SPAIN 25-34 35-44 100% Share your point of view 34 Challenge your point of view 45-54 No point of view Share point 31 55+ Challenge your point of view of view No point of view 26 75% 24 Share your point of view 51 56 58 50% 49 52 43 No point 57 33 of view 2020 25% 28 62 66 14 10 9 0% 17 16TH TV news Social media Printed newspapers 14 BRAZIL Challenge 12 point of view 11 JUNE 100% Share your point of view Challenge your point of view No point of view 10 Challenge your point of view No0% point of view 25% 50% 75% 75% Q5c_2020.TIME) Thinking about the different kinds of news available to you, do you prefer getting news from sourcesShare your that share/challenge/havepoint of view no point of view. Base: 18–24 = 8639, 25–34= 12603, 35–44 = 12643, 45–54 = 11605, 55+ = 23773. 50% 52 47 48 49 43 41 In reality this is not a zero-sum game. Most people like to mix news that they can trust with a range of opinions that challenge 25% (BRITISHor support their existing views. We do find, however, that those with extreme political views are significantly less attracted to objective news – and these are often the same people that are 10 5 5 increasingly distrustful of mainstream media. Younger people are 0% TV news Social media Printed newspapers more ambivalent. They still turn to trusted mainstream brands at 00.01 times of crisis but they also want authentic and powerful stories, Q5c_2020. Thinking about the different kinds of news available to you, do you prefer getting and are less likely to be convinced by ‘he said, she said’ debates news from sources that share/challenge/have no point of view. Base: Total sample (excluding don’t knows): Spain = 1756, Brazil = 1901. that involve false equivalence.

Interpreting notions of objectivity is not an easy task at a time of MISINFORMATION AND DISINFORMATION rising political polarisation, even on issues in which the weight of scientific evidence overwhelmingly favours one side. Public media More than half (56%) of our sample across 40 countries remains organisations like the BBC in the UK have faced criticism for adopting concerned about what is real and fake on the internet when a ‘he said, she said’ approach to coverage of topics such as climate it comes to news. Concern tends to be highest in parts of the change as theyEMBARGOED: seek to present both sides of arguments. Some UK Global South such as Brazil (84%), Kenya (76%), and South Africa news organisations have taken a different approach to these issues, (72%) where social media use is high and traditional institutions with the Guardian adopting the term ‘climate emergency’ to describe are often weaker. Lowest levels of concern are in less polarised the urgency of the story. Our recent qualitative study of news European countries like the Netherlands, Germany, and Denmark. behaviours amongst under-35s showed that younger age groups in particular tend to respond well to approaches and treatments The biggest increase in concern came in Hong Kong this year (+6) that take a clear point of view (Flamingo 2019). Our survey data as the conflict between the government and student protesters also show that, across countries, young people are also less likely continued and also in Finland (+4), where we see higher than to favour news with no point of view. average concern over false and misleading information from foreign governments. REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS 18 / 19 Slide 23

PROPORTION CONCERNED ABOUT WHAT IS REAL AND WHAT IS FAKE ON THE INTERNET WHEN IT COMES TO NEWS – ALL MARKETS

USA UK Mexico Finland Hong Kong 56% 100% Concern high Brexit concerns Post-election Fears of Violent worry about ahead of election ease concerns Russian street ease influence protests what is real or fake 84 (no change). -7 75% 76 76 -8 Average of 72 40 markets. 67 +4 65 65 65 65 64 63 63 63 62 62 62 60 60 60 +6 58 57 56 50% 55 54 54 51 51 50 49 49 46 45 45 45 42 40 37 37 35 32 25%

0% ZAKENPORBRA ZAKENPORBRA USA SGN CAN SPA CHL AUS UK GRE MYS TURIREFRA ARG MEXKOR ROU FINPHL CRO JPN ITA BGRHK BELHUN SWE POLSUITWNCZE 2020NOR GERDENAUT NLDSVK Q_FAKE_NEWS_1. Please indicate your level of agreement with the following statement. Thinking about online news, I am concerned about what is real and what is fake on the internet. Base: Total sample in each market ≈ 2000, Taiwan = 1027.

Domestic politicians are seen as most responsible (40%) for But political starting positions can16TH make a big difference when false and misleading information online, followed by political it comes to assigning responsibility for misinformation. In activists (14%), journalists (13%), ordinary people (13%), and the most polarised countries, this effectively means picking foreign governments (10%). Despite widespread media coverage your side. Left-leaning opponents of Donald Trump and Boris of alleged attempts by outside powers to undermine elections, Johnson are far moreJUNE likely to blame these politicians for REUTERSit is striking INSTITUTE that it FOR is the THE rhetoric STUDY OF and JOURNALISM behaviour / FURTHER of national ANALYSIS spreading lies and half-truths online, while their right-leaning politicians that is considered the biggest problem. This echoes supporters are more likely to blame journalists. In the United Slidearguments 24 from scholars that misinformation often comes from States more than four in ten (43%) of those on the right blame the top (and not from ordinary people). journalistsTIME) for misinformation – echoing the President’s anti- media rhetoric – compared with just 35% of this group saying they are most concerned about the behaviour of politicians. PROPORTION THAT SAY THEY ARE MOST CONCERNED For people on the left the situation is reversed, with half (49%) ABOUT FALSE OR MISLEADING INFORMATION FROM blaming politicians and just 9% blaming journalists. EACH OF THE FOLLOWING – ALL MARKETS REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS By contrast, in the Netherlands, there is less political polarisation 40% politicians (BRITISHSlideand less picking 25 of sides required, though those on the right are More in... USA, Brazil, Philippines, South Africa still twice as critical of the role of journalists.

activists 14% PROPORTION THAT SAY THEY ARE MOST CONCERNED More in... Hong Kong, Norway, Czech00.01 Republic ABOUT FALSE OR MISLEADING INFORMATION FROM EACH OF THE FOLLOWING BY LEFT-RIGHT POLITICAL LEANING – journalists SELECTED COUNTRIES 13% Left Right More in... Greece, South Korea, Croatia, USA 75% Right 13% ordinary people More in... Japan, Taiwan, South Korea UK USA NETHERLANDS 61 Le

50% 10% foreign governments 49 MoreEMBARGOED: in... Denmark, Finland, Germany, Netherlands 43

Q_FAKE_NEWS_2020b. Which of the following, if any, are you most concerned about online? 35 False or misleading information from… Base: Total sample = 80155. 25% 27 28 28

We see a slightly different pattern in parts of Asia where 17 the behaviour of ordinary people – for example, in sharing 11 11 9 false information without considering the consequences – is 5 0% considered the biggest problem in Japan (27%) and is second only Domestic Journalists Domestic Journalists Domestic Journalists to domestic politicians in Taiwan (29%). politicians/ and news politicians/ and news politicians/ and news parties organisations parties organisations parties organisations

Q_FAKE_NEWS_2020b. Which of the following, if any, are you most concerned about online? False or misleading information from… Base: Left/Right: UK = 352/283, USA = 507/458, Netherlands = 334/305. REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS Slide 26

Reuters Institute for the Study of Journalism / Digital News Report 2020 REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS 20 Slide 28

CHANNELS OF MISINFORMATION MOST CONCERNING PLATFORM FOR FALSE AND MISLEADING INFORMATION – SELECTED COUNTRIES It is perhaps no surprise that people see social media as the biggest source of concern about misinformation (40%), well 50% Facebook YouTube Twitter WhatsApp 47 47 6 Philippines ahead of news sites (20%), messaging apps like WhatsApp (14%), 2 and search engines such as Google (10%). 15 35 5 35 35 USA 7 PROPORTION THAT SAY THEY ARE MOST CONCERNED ABOUT 32 31 7 FALSE OR MISLEADING INFORMATION FROM EACH OF THE 29 FOLLOWING – ALL MARKETS 10 25% 25 31 24 Korea 4 7 6 15 12 Japan 12 32 10 9 9 7 7 7 7 7 40% 20% 14% 10% 6 5 6 2020 29 Social News sites/ Messenger Search eg 4 4 3 4 2 Kenya media apps apps Google 0% 7 Philippines USA South Korea Japan Kenya Brazil 25 Q_FAKE_NEWS_2020c. Which of the following, if any, are you most concerned about online? 24 False or misleading information from… Base: Total sample = 80155. 7 FAKE_NEWS_2020c. Which of the following, if any,16TH are you most concerned about online? Please Brazil select one. False or misleading information from… Base: Total sample in each market ≈ 2000. 3 35 0% 25% Breaking the social data down further, across all countries 29% say they are most concerned about Facebook, followed The coronavirus crisisJUNE has reminded us that these networks can by YouTube (6%) and Twitter (5%). But in parts of the Global be used to spread all kinds of damaging misinformation, not just South, such as Brazil, people say they are more concerned about about politics. A range of unsubstantiated conspiracy theories closed messaging apps like WhatsApp (35%). The same is true in have been doing the rounds including one linking COVID-19 Chile, Mexico, Malaysia, and Singapore. This is a particular worry to 5G networks and another suggesting that the virus was a because false information tends to be less visible and can be biologicalTIME) weapon originating in a Chinese military research harder to counter in these private and encrypted networks. facility – though it is important to note that politicians and the By contrast, in the Philippines (47%) and the United States media have also played a role in bringing these ideas to a wider (35%) the overwhelming concern is about Facebook, with other public. In our April survey, almost four in ten (37%) said they REUTERSnetworks INSTITUTE playing aFOR minor THE STUDYrole. Twitter OF JOURNALISM is seen to/ FURTHER be the biggest ANALYSIS had come across a lot or a great deal of misinformation about Slideproblem in 27 Japan and YouTube in South Korea. Facebook is used COVID-19 in social media like Facebook and Twitter, and 32% much less widely in both of these countries. (BRITISHvia messaging apps like WhatsApp.

MOST CONCERNING PLATFORM FOR FALSE AND MISLEADING INFORMATION – ALL MARKETS 00.01

32% Most concerned 35% 31% about Twitter Most concerned Most concerned in Japan about Facebook about YouTube in USA (just 7% in South Korea WhatsApp)EMBARGOED: 47% Most concerned 33% about Facebook in the Philippines Most concerned 35% about WhatsApp Most concerned in Malaysia about WhatsApp in Brazil 30% Facebook Most concerned about Facebook WhatsApp/other messaging in South Africa Twitter 23% WhatsApp YouTube

FAKE_NEWS_2020c. Which of the following, if any, are you most concerned about online? Please select one. False or misleading information from… Base: Total sample in each market ≈ 2000, Taiwan = 1027. REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS 20 / 21 Slide 29

PROPORTION THAT SAY THEY HAVE SEEN FALSE OR POLITICAL ADS AND FACEBOOK MISLEADING INFORMATION ABOUT CORONAVIRUS ON EACH (APRIL 2020) – AVERAGE OF SIX COUNTRIES One flashpoint is likely to be the extent of any intervention by social networks in the run-up to the US presidential election Social media (eg Facebook, Twitter) 37% 30% 15% 18% in November. A burning question is the extent to which social Messaging apps (eg WhatsApp, networks should remove misleading advertising from politicians 32% 29% 19% 20% Facebok Messenger) standing for election. Facebook has exempted candidates from Video sites normal fact-checking processes but has increased transparency 26% 30% 20% 24% (YouTube, Vimeo) so it is clear what campaigns are being run and how much is Search engines being spent. Twitter has taken a different view and has restricted (Google, Bing, Naver) 19% 35% 26% 20% political ads on its platform entirely. A lot/Great deal Some Little/None Don’t know In our January survey we find that 58% of our respondents Q13 (Apr. 2020). How much false or misleading information about coronavirus (COVID-19), if any, think that misleading ads should be blocked by social media do you think you have seen from each of the following in the last week? Base: Total sample: USA, UK, Germany, Spain, Argentina, South Korea = 8522. companies, with just a quarter (26%) thinking it should not be up to technology companies to decide what2020 is true in this REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS context. Support for blocking such ads was strongest in countries Given these concerns, Facebook has stepped up funding like Germany, France, and the UK where political advertising is Slidefor independent 30 fact-checkers and a number of platforms already tightly controlled, and was weakest in countries that have including Facebook, Twitter, and YouTube have taken down traditionally worried about regulating free speech – such as the misinformation that breached guidelines, including a video 16TH United States and the Philippines. from Brazilian President Jair Bolsonaro.7 By contrast people are less keen on media companies omitting Digital literacy has also been a priority, with a prominent link potentially misleading statements by candidates in the course of to trusted information sources pinned at the top of Facebook a campaign. Over halfJUNE (52%) say that news organisations should news feeds in many countries. Algorithms have been tweaked to report statements prominently because it is important to know prioritise official and trusted news sources. what the politicians said. This is consistent with our earlier data It may be that COVID-19 finally supercharges the fight against which suggest the majority of people would like to make up their misinformation and gives social networks more confidence to own mindTIME) rather than be told what to think by a journalist – or to take down damaging or dubious content, but these judgements feel that information was being withheld. are likely to get more difficult once the immediate crisis is For further analysis see section 2.3: How do People Want the Media to over, the blame game begins, and the cut and thrust of normal Cover Politics? politics resumes. (BRITISH

00.01

EMBARGOED:

7 https://news.sky.com/story/coronavirus-brazil-president-refuses-to-ramp-up-covid-19-lockdown-as-facebook-pulls-video-11966279 Reuters Institute for the Study of Journalism / Digital News Report 2020 REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS 22 Slide 32

CHANGING BUSINESS MODELS AND THE GROWTH PROPORTION THAT PAID FOR ANY ONLINE NEWS IN LAST OF PAID CONTENT YEAR (2014–20) – SELECTED COUNTRIES USA 9 country average It is too early to predict the full impact of the COVID-19 crisis on the 25% news industry but it is almost certain to be a catalyst for more cost- 1. Political boost 2. Newspapers tighten 9 country average for liberal media paywalls, better target cutting, consolidation, and even faster changes in business models. (Trump Bump) potentials 20% 20% Germany While some publications report growth in digital subscriptions, some publishers say advertising revenue has fallen by up to 50% and 16% USA 15% many newspapers have cut back or stopped printing physical copies 13% 8 11% and laid off staff. In Australia, News Corporation suspended print 10% 10% production of around 60 newspapers, while in the UK analysts warn 9% that as many as a third of journalists in the media could lose their 5% job as a result of the pandemic.9 All of this puts more focus on reader payment models online – including subscription, membership, 0% 2014 2015 2016 2017 2018 2019 2020 donation, and micropayment – and on the issue of trust which

underpins these. Q7a. Have you paid for ONLINE news content, or accessed a paid-for2020 ONLINE news service in the last year? Total 2014–20 samples ≈ 2000. Note: 9 country average includes USA, UK, France, Spain, Italy, In the last 12 months more publishers have started charging for Germany, Denmark, Japan, and Finland. content or tightening paywalls and this is beginning to have an impact. Across countries we have seen significant increases in In the wake of the pandemic, the 16THNew York Times and the Atlantic the percentage paying for online news, including a jump of four are among US publications reporting substantial increases in percentage points in the United States to 20% and eight points digital subscriptions while the Guardian has seen a boost to the in Norway to 42%. We’ve also seen increases in Portugal, the numbers of contributors. Some publishers have emphasised Netherlands, and Argentina, with the average payment level the value of trusted and accurate journalism through a series of coronavirus-themedJUNE messaging campaigns designed to increase also up in nine countries that we have been tracking since 2013. REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS subscriptions or donations. But the crisis has alsoSlide 33raised new Looking back, we can detect two clear waves of subscription dilemmas around paywalls, with many growth in the United States. The first was sparked by the election news organisations like the New York Times of Donald Trump, with many younger and liberal voters looking and El TIME)Pais in Spain dropping their paywalls to support publications that could hold the President to account. for a time. Others like the Financial Times ThisREUTERS year’s INSTITUTE uptick mayFOR THE be partly STUDY drivenOF JOURNALISM by a new / FURTHER election ANALYSIS cycle but have offered a subset of content for free to Slidealso by publisher 31 tactics to restrict the amount of content people ensure that critical public health content can see for free combined with special deals. was available to all. (BRITISH PROPORTION THAT PAID FOR ANY ONLINE NEWS IN LAST YEAR – SELECTED COUNTRIES

00.01

Norway 42% (+8) 26% 14% Sweden 27% (-) Average in Nordic Netherlands countries pay (+3) Finland 19% (+3) 20% Denmark 17% (+2) USA (+4) EMBARGOED:

14% ALSO: Australia Germany 10% France 10% Argentina 11% Spain 12% Portugal 10% Japan 8% Austria 11% Poland 20% UK 7%

Q7a. Have you paid for ONLINE news content, or accessed a paid-for ONLINE news service in the last year?Base: Total sample in each market ≈ 2000.

8 https://www.poynter.org/business-work/2020/a-qa-with-tampa-bay-times-chairman-and-ceo-paul-tash-about-the-times-print-reduction 9 https://www.endersanalysis.com/reports/enders-analysis-calls-government-support-news-magazine-media 22 / 23

Many fear growing levels of information inequality, where people In all three countries the majority of those who pay only with less money become more dependent on social media and subscribe to one title, but in Norway more than a third (38%) other low-quality news while those who can afford it get better have signed up for two or more – often a national title and a information. Currently levels of concern about this issue from local newspaper. It is a similar story in the United States with respondents are highest in the United States (24%) and Norway 37% taking out two or more subscriptions, with additional (17%). Lower concern in the United Kingdom (9%) may be payment often for a magazine or specialist publication such as REUTERSbecause INSTITUTE of widespread FOR THE availability STUDY OF JOURNALISMof high-quality / FURTHER free news ANALYSIS from the New Yorker, the Atlantic or The Athletic (sport). By contrast, Slidethe BBC News 34 website, many widely read popular newspapers in the UK a second or third subscription is rare, with three- and digital-born titles, and titles like the Guardian, with its open quarters (74%) paying for just one title. donation model. In the UK and Norway, three-quarters (75%) pay for subscriptions with their own money, around 5% are on a free trial, with the PROPORTION CONCERNED THAT OTHERS MAY BE MISSING rest paid for by someone else (by work or a gift). The free trials OUT ON NEWS BECAUSE OF PAYWALLS – USA, NORWAY, UK seem to be twice as prevalent in the United States (10%), where competition for a limited pool of subscribers is extremely fierce. USA 24% 23% 42% 2020 For further results from our detailed pay survey see section 2.1: Norway 17% 29% 45% How and Why People are Paying for Online News

UK 9% 22% 58% Donations for news are relatively new, though Wikipedia and National Public Radio have generated16TH much of their income Concerned Neither Don’t know Not concerned this way for many years. Our survey showed a growing array of publications to which people are prepared to give money. P3_2020_1. Some news providers now charge money for access to their online news content. In the United States 4% now say they donate money to a news How concerned are you, if at all, that other people might be missing out on news from sources that they have to pay for? Base: USA = 2055, UK = 2011, Norway = 2010. organisation, 3% in Norway,JUNE and 1% in the UK. The Guardian has one of the most successful donation models of any major brand, with over a million people having contributed in the last WINNER-TAKES-MOST IN SUBSCRIPTIONS year.10 According to our data, almost half of all the relatively small numberTIME) of donations in the UK (42%) go to the Guardian, This year we conducted a second more detailed survey in three REUTERSbut most INSTITUTE are one-off, FOR THE with STUDY an average OF JOURNALISM payment / FURTHER of less ANALYSISthan £15. countries (USA, UK, and Norway) to understand more about Other beneficiaries include small community websites such as patterns and attitudes to online news payment. This research Slidethe Bristol 36 Cable, YouTube channels, and podcasts such as the clearly shows that a few large national news brands have been one produced by the investigative co-operative Bellingcat. the biggest winners in all three countries. Around half of those that subscribe to any online or combined package in the United States use the New York Times or the Washington Post and a(BRITISH similar proportion subscribe to either The Times or the Telegraph in the UK, though in much smaller numbers. In Norway the quality newspaper Aftenposten leads the field (24% of those that pay), along with two tabloid newspapers that operate premium paywall models, VG (24%) and Dagbladet (14%).00.01 One surprise in these data is the extent to which people subscribe to local and regional newspapers online in Norway (64%). Most local news publishers in Norway charge for online news, with a total of 129 REUTERSdifferent INSTITUTE local titles FOR THEmentioned STUDY OF by JOURNALISM our respondents. / FURTHER In the ANALYSIS United States 30% subscribe to one or more local titles, with 131 different Slidetitles mentioned. 35 By contrast in the UK only a handful of local publishers have put up a paywall.

PROPORTION WITH ACCESS TO PAID NEWS THAT HAVE ACCESS TO EACH BRANDEMBARGOED: – UK, USA, NORWAY In looking through the donation lists, it is striking to see so many political partisan websites such as The Daily Kos and UK USA Norway Patriot Post in the US and The Canary and Novara media The Times New York Times Local Papers in the UK. In Norway, a significant number of donations to 39% 39% LP 64% news services go to one of three far-right websites, Resett, Telegraph Washington Post Aftenposten Document, and HRS. With advertisers increasingly wary of 20% 31% 24% controversial political content, donations are proving one Local Papers Local Papers Verdens Gang important way to tap into the loyalty of committed partisans. LP 5% LP 30% 24%

PAY3. You said that you currently have access to a single news brand. Please list all of the online brands you’d normally have to pay for that you currently have access to. Base: All who have access to a single news brand: USA = 584, UK = 256, Norway = 751.

10 1.16m have contributed, including 821,000 ongoing monthly donations (and subscriptions). https://www.theguardian.com/media/2020/apr/29/guardian-reports-surge-in-readers- support-over-past-year ReutersREUTERS Institute INSTITUTE for the FOR Study THE of Journalism STUDY OF JOURNALISM/ Digital News / FURTHER Report 2020 ANALYSIS 24 Slide 37

PROPORTION OF DONATERS THAT DONATE TO EACH BRAND In federated systems like Germany and the United States local – UK, USA, NORWAY newspapers play a critical role in providing accountability and this is reflected in these scores. The value that people place UK USA Norway on their local newspaper is closely correlated with the higher subscription and donation rates that we see for local news in Guardian Guardian Resett Norway and the United States when compared to the UK. 42% 13% 13% Wikipedia National Public Document For further analysis see section 2.4: Global Turmoil in the Neighbourhood: Radio 13% 11% 7% Problems Mount for Regional and Local News

DONATE2. You say you currently donate to a news brand. Which brands do you donate to? Base: All those who have donated: UK = 197, US = 433, Norway = 211. GATEWAYS AND INTERMEDIARIES

After the 2008 economic crash many publishers saw advertising In terms of access points for online news, habits continue to revenues drop by around a third and many worry that the become more distributed – as more and more people embrace aftermath of the COVID-19 pandemic could be even worse. various digital platforms that were initially used2020 most intensely Online subscriptions and donations may make up for some of by younger people. Across all countries, just over a quarter (28%) this, especially at a time when trusted news matters more than prefer to start their news journeys with a website or app, followed ever, but for most publishers any extra reader revenue won’t be by social media (26%) – up two percentage points on last year. nearly sufficient. Industry expectations need to be realistic as Once again, though, we see very different16TH patterns with 18–24s, many people are likely to have less disposable income and the the so-called Generation Z. This group has a much weaker direct majority of non-payers are still broadly happy with free sources. connection with news brands (16%) and is almost twice as likely to prefer to access news via social media (38%).

UNCERTAIN FUTURE FOR LOCAL NEWS If younger groups cannotJUNE be persuaded to come to specific websites50% and apps, publishers may need to focus more on how The COVID-19 disruption is likely to hit local news providers to build audiences through third-party platforms like Facebook, hardest, given their continuing dependence on both print and Twitter, YouTube, and Snapchat. This may become more attractive digital advertising. Commercial local newspapers in particular REUTERSfor publishers INSTITUTE-1 now FOR Facebook THE STUDY has OF started JOURNALISM to offer / FURTHER direct ANALYSISpayments TIME)+2 are a key component of democracy in many countries, employing for its dedicated ‘news tab’, but aggregated environments have Slide25% 28 39 the majority of journalists, highlighting local issues, and holding not yet proved a good26 environment25 to build the loyalty and politicians to account. In Denmark, Sweden, and some other attribution that will be needed for long-term relationships. REUTERScountries INSTITUTE the government FOR THE STUDY has stepped OF JOURNALISM in to provide / FURTHER subsidies ANALYSIS and short-term relief for the sector.11 Without more support further MAIN GATEWAYS TO NEWS – ALL MARKETS8 7 5 Slideclosures and 38 cuts seem inevitable. But how much would people 0% Direct Social Search Mobile Aggregators Email 50% miss their local newspapers if they were no longer there? (BRITISH media ALL AGESalerts

-1 72% PROPORTION OF READERS WHO WOULD MISS THEIR LOCAL +2 side door NEWSPAPER A LOT – SELECTED COUNTRIES access 00.01 25% 28 26 25

54% 49% 39% 25% 8 7 Germany Norway USA UK 5 0% Direct Social Search Mobile Aggregators Email media alerts

L4_2020. How much would you miss your local newspaper/website if it went out of business? 50% GEN Z Showing code ‘a lot’. Base: Those who consume local newspapers in last week. Germany = 1136, +2 Norway = 1283, USA = 730, UK = 835. 84% EMBARGOED: 38 side door access Our research shows that local newspapers (and their websites) 25% -1 25 are valued much more in some countries than others. Over half of those who regularly read local newspapers in federal 16 +2 Germany (54%) say they would miss them ‘a lot’ if they were no 9 8 longer there, 49% in Norway, and 39% in the United States. By 4 PROPORTION OF READERS WHO WOULD MISS THEIR LOCAL 0% contrast only 25% of those who regularly read newspapers in Direct Social Search Mobile Aggregators Email NEWSPAPER A LOT – SELECTED COUNTRIES media alerts the UK, 18% in Argentina, and 13% in Taiwan say the same. The

value placed on local news seems to be partly related to the Q10a. Which of these was the MAIN way in which you came across news in the last week? importance that countries place on their regions more generally Base: All/18-24s that came across online news in the last week: All markets = 74181/9083. – and the extent to which local politics matters. 54% 49% Germany Norway

11 https://euobserver.com/coronavirus/147956

39% 25% USA UK

L4_2020. How much would you miss your local newspaper/website if it went out of business? Showing code ‘a lot’. Base: Those who consume local newspapers in last week. Germany = 1136, Norway = 1283, USA = 730, UK = 835. 24 / 25

MOBILE AGGREGATORS PUBLISHERS FIGHT BACK WITH EMAIL AND OTHER LOYALTY TACTICS Some publishers have been reporting significant increases in referrals from mobile apps such as Apple News, Upday, and various Faced with the growing power of platforms, publishers have been Google products including Google News and Google Discover. working hard to build direct connections with consumers via email, Some of these (Apple News and Google) are very hard to measure mobile alerts, and podcasts. Across countries, around one in six in surveys because respondents often cannot recall these services (16%) access news each week via email, with most of these (60%) – especially if they access via a notification. Apple News is still only accessing a briefing of general or political news, often sent in the available in a few countries but is accessed by 29% of iPhone users morning. But publishers have been extending the range of formats, in the United States and 22% in the UK. Our data suggest that the increasingly offering ‘pop up emails’ on subjects like coronavirus premium Apple News + ($9.99 a month) has not yet had much and the 2020 presidential elections. Emails have proved effective impact outside the US, where the service includes content from in attracting potential new subscribers, as well as encouraging the Wall Street Journal and the Los Angeles Times. existing users to come back more frequently.

40% Aggregators remain more important in some countries (South Despite thisUK -revival Apple Smartphone in the newsletter format we find surprising Korea, Japan) than others but overall access remains constant differences in the level of adoption across countries. In the United UK 2020 at around 7% and there has been little movement in the States one in five (21%) access a news email weekly, and for almost headline numbers for specific apps. Line News reaches 22% and half of theseUSA it- Apple is their Smartphone primary way of accessing news. Northern Gunosy20% 12% in Japan each week, while Kakao Channel (25%) EuropeanUSA countries have been much slower to adopt email news is a popular mobile aggregator in South Korea. TopBuzz, a new channels, with only 10% using email16TH news in Finland and 9% in the UK. REUTERSartificial INSTITUTE intelligence FOR THE(AI) STUDY driven OF aggregator JOURNALISM /which FURTHER often ANALYSIS surfaces low-quality news sources, has attracted 5% usage in the United The New York Times now offers almost 70 different scheduled SlideStates and 40 is also popular in countries in the Global South such emails, with the popular morning briefing now reaching 17 million as0% the Philippines (9%), Kenya (11%), and Brazil (9%). subscribers. The size and importance of this audience has meant 2016 2017 2018 2019 2020 JUNE an increased profile for flagship emails, with senior journalists appointed as hosts to help guide users through the news each day.12 PROPORTION OF iPHONE USERS THAT USE APPLE NEWS (2016–20) – USA AND UK Americans get, on average, more emails from different news providers (4) compared with Australians (3) or British (3). USA USA – Apple iPhone TIME) Across countries almost half (44%) say they read most of their UK UK – Apple iPhone news emails. Emails tend to be most popular with older age groups, 40% while mobile notifications are more popular with the young and continue to grow in many countries. REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS 29% SlideFor further 42analysis see section 2.2: The Resurgence and Importance 22% (BRITISHof25% Email Newsletters 20% 14% 11% 11% 8% PROPORTION THAT USED EACH AS A WAY OF GETTING NEWS 4% IN THE LAST WEEK (2015–20) – SELECTED COUNTRIES 3% 00.01 0% 2016 2017 2018 2019 2020 UK USA France Spain Finland REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS 0% Email newsletter Q10c. When using the internet for news, have you used any of the following sites or mobile apps 25% 25%2015 2016 2017 2018 2019 2020 Slidethat aggregate different41 news links in the last week? Base: All/Apple iPhone users 2016–20: USA ≈ 2055/530, UK ≈ 2011/583. 21% 21%

15% 14% 15% 10% 10% PROPORTION THAT USED EACH MOBILE AGGREGATOR 9% 9% IN THE LAST WEEK – SELECTED COUNTRIES

0% EMBARGOED: 2015 2016 2017 2018 2019 2020

Google News Apple News Upday Flipboard 25% Mobile notifications 22% USA 17% 11% 3% 4% 18% UK 6% 8% 2% 2% 14% 16% Germany 14% 5% 5% 2% 13% 13% 10% 9% France 20% 5% 3% 2% 8% 7% Spain n/a 7% 6% 5% 0% Japan 21% 5% 2% 1% 2015 2016 2017 2018 2019 2020 All markets 23% 7% 5% 3% Q10. Thinking about how you got news online (via computer, mobile or any device) in the last Q10c. When using the internet for news, have you used any of the following sites or mobile apps week, which were the ways in which you came across news stories? Base: Total sample in each that aggregate different news links in the last week? Base: Total sample in each market ≈ 2000. market 2015-20 ≈ 2000.

12 https://www.niemanlab.org/2020/04/the-new-york-times-morning-email-newsletter-is-getting-an-official-host-and-anchor REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS Slide 43

Reuters Institute for the Study of Journalism / Digital News Report 2020 26

PODCASTS AND THE RISE OF AUDIO Podcast listeners tend to be younger (see next chart) and mainly REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS listen via headphones/mobile phones. These demographics and In the last few years podcasts have become another important habits are not just of interest to subscription businesses but Slide 46 channel for driving loyalty to specific news brands. The Daily from also to broadcasters who, as we’ve already noted, are finding it the New York Times attracts 2 million listeners a day and, although harder to reach under-35s through linear programming. In the the advertising revenue is substantial, the main strategic aim REUTERSUK half ofINSTITUTE all podcasts FOR THE are STUDY listened OF JOURNALISM to by under-35s, / FURTHER despite ANALYSIS PROPORTION THAT USED EACH TYPE OF PODCAST is to attract new subscribers and to build habit with existing these making up just a third of our sample. By contrast, the Slide 45 IN THE LAST MONTH BY AGE – USA ones (Newman and Gallo 2019). The Guardian (UK), Aftenposten majority of those listening to traditional news bulletins and (Norway), and Les Echos (France) are amongst publishers to programmes on the radio each week are over 50. News and 23 have launched successful daily news podcasts in the last two current affairs 34 18-24s years. During the coronavirus lockdowns some podcast listening PROPORTION THAT USED A PODCAST IN THE LAST MONTH 21 Lower interest reportedly fell by up to 20% – a reminder of how integral AND RADIO NEWS IN THE LAST WEEK BY AGE – UK in news-based podcasts have become to commuting habits and other activities 13 podcasts with outside the home.13 At the same time, listening to other types of 18-24s when 13 compared podcasting has shifted to the home and there have been some with young breakout hits, including Das Coronavirus-Update, a 30-minute 2020 millennials show featuring one of Germany’s top virologists, which reached Specialist 18 Podcast Radio news (health, tech, No. 1 in the podcast charts there. listeners by age listeners by age business) 28 Proportion of each age group Proportion of each age group 19 listening monthly listening weekly 50% 16TH 7

43 8 40 38 33 32 Lifestyle 23 25% JUNE (food, fashion, 22 celebs) 23 19 19 16 10 11 9 0% TIME) 4 18-24 25-34 35-44 55+45-54 18-24 25-34 35-44 55+45-54

Q11F. A podcast is an episodic series of digital audio files, which you can download, subscribe, Sports 12 or listen to. Which of the following types of podcast have you listened to in the last month? 100% Q3. Which, if any, of the following have you used in the last week as a source of news? 12 Base: 18–24/25–34/35–44/45–54/55+ = 201/282/343/349/836. 84 11 75% 76 76 6 72 67 (BRITISH The underlying picture65 remains 65 65 65 one64 63of growth. 63 63 62 Our 62 data 62 show Many podcasts contain an informational element (sport, lifestyle, 2 60 60 60 58 an overall rise in podcast listening to 31% (+3) across a basket 57true56 crime)55 54 but 54 podcasts specifically about news and politics are 50% 51 51 50 49 49 46 of 20 countries we have been tracking since 2018. Almost four amongst the most widely listened to. 45About 45 45half42 of podcast users 40 37 37 True crime 18 18-24 in ten access monthly in Spain (41%), Ireland (40%), Sweden listen to a news podcast in the US, where the market has developed35 32 25%(36%), Norway (36%), and the United States00.01 (36%). By contrast furthest. Podcast users in the United States say that the format 20 25-34 REUTERSusage in INSTITUTE the Netherlands FOR THE STUDY (26%), OF GermanyJOURNALISM (24%), / FURTHER and ANALYSISthe UK gives greater depth and understanding of complex issues (59%) 16 35-44 0%(22%) is nearer to a quarter. and a wider range of perspectives (57%) than other types of media. 6 45-54 Slide 44ZAKENPORBRA USA SGN CAN SPA CHL AUS UK GRE MYS TURIREFRA ARG MEXKOR ROU FINPHL CRO JPN ITA HK BRG BELHUN SWE POLSUITWNCZE NOR GERDENAUT NLDSVK 6 55+

0% 25% 50% PROPORTION THAT USED A PODCAST IN THE LAST MONTH – SELECTED COUNTRIES Q11F. A podcast is an episodic series of digital audio files, which you can download, subscribe 50% or listen to. Which of the following types of podcast have you listened to in the last month? Base: 18–24/25–34/35–44/45–54/55+ = 258/378/348/243/828. +2 +3 +5 +3 41 40EMBARGOED:+4 +5 36 36 36 +3 34 +4 33 32 32 32 +3 +5 30 +3 29 28 28 25% 26 26 26 24 24 31% 22 access a podcast monthly, up from 29% in 2019

0% SPA IRE SWE NOR ARGUSA CAN AUS SUI ITA FINSGN UKGERJPNFRANLDBELDENAUT

Q11F. A podcast is an episodic series of digital audio files, which you can download, subscribe, or listen to. Which of the following types of podcast have you listened to in the last month? Please select all that apply. Total sample in each market ≈ 2000. Note: We excluded markets with more urban samples as well as those where we are not confident that podcasts is a term sufficiently well understood to produce reliable data.

13 https://podnews.net/article/coronavirus-covid19-affecting-podcasting 26 / 27

REUTERSNews podcasts INSTITUTE are FOR most THE popular STUDY OF with JOURNALISM 25–34s (young / FURTHER millennials). ANALYSIS WHICH PLATFORMS DO PEOPLE USE TO ACCESS Those aged 18–24 are less likely to listen to news podcast but PODCASTS? Slideare some of 46 the heaviest consumers of lifestyle and celebrity podcasts as well as true crime. Podcasts have traditionally been associated with Apple devices, but that is changing rapidly. In the last 18 months Spotify has invested over $500m in podcasting and has reported a doubling PROPORTION THAT USED EACH TYPE OF PODCAST of podcast listens.14 Broadcaster apps like BBC Sounds, ABC IN THE LAST MONTH BY AGE – USA Listen (Australia), NPR One (USA), and SR Play (Sweden) now offer original podcasts in addition to live and catch-up radio. News and 23 current affairs Google has started to promote podcasts within search and has 34 18-24s revamped its own podcast service. Our data from the UK show 21 Lower interest in news-based BBC Sounds (28%) is similar to Apple (26%) and Spotify (24%) in 13 podcasts with terms of access, while Spotify (25%) is now ahead of Apple (20%) 18-24s when 13 compared and Google (16%) in the United States. The US has a much larger with young ecosystem of widely used smaller podcast apps and services millennials including TuneIn, Podcast Addict, and Stitcher.2020 Spotify (40%) is REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS Specialist 18 ahead in Sweden, the country of its birth and also in Australia (health, tech, business) 28 Slide(33%). In Sweden 47 the national radio broadcaster reaches more 19 than a quarter of podcast users. 7 16TH 8 PROPORTION OF PODCAST LISTENERS THAT USE EACH PLATFORM – SELECTED COUNTRIES Lifestyle 23 JUNE (food, fashion, UK celebs) 23 16 BBC Sounds 28 9 Apple podcasts 26 4 TIME)Spotify 24 Google podcasts 10 Website (stream) 6 Sports 12 Audible 5 12 Podcast Addict 5 11 Castbox 3 6 (BRITISH TuneIn 3 2 Pocket Casts 2 Podbean 2

True crime 18 18-24 Overcast 1 1 Public broadcaster app 20 00.01 25-34 Stitcher 0% 25% 50% 16 35-44 6 45-54 USA 6 55+ Spotify 25 Apple podcasts 20 0% 25% 50% Google podcasts 16 15 Q11F. A podcast is an episodic series of digital audio files, which you can download, subscribe Pandora or listen to. Which of the following types of podcast have you listened to in the last month? Website (stream) 11 Base: 18–24/25–34/35–44/45–54/55+ = 258/378/348/243/828. Audible 10 EMBARGOED: NPR One 9 The deep connection that many podcasts seem to create could TuneIn 8 7 be opening up opportunities for paid podcasts, alongside Deezer 7 advertising-driven models. Almost four in ten Australians (39%) Podcast Addict Overcast 6 said they would be prepared to pay for podcasts they liked, 38% Pocket Casts 6 in the United States, and a similar number in Canada (37%). RadioPublic 6 Willingness to pay was lower in Sweden (24%) and the UK (21%) Stitcher 6 Non-profit broadcaster app where so many popular podcasts come from free-to-air public Castbox 5 broadcasters. 0% 25% 50%

POD2. Which of the following apps or websites do you mainly use to find and play podcasts? Base: All that listened to podcasts in the last month: UK = 440, USA = 746.

14 https://www.theverge.com/2020/2/5/21123905/spotify-earnings-q4-2020-podcasting-investment-operating-loss ReutersREUTERS Institute INSTITUTE for the FOR Study THE of Journalism STUDY OF JOURNALISM/ Digital News / FURTHER Report 2020 ANALYSIS 28 Slide 48

PROPORTION OF PODCAST LISTENERS THAT USE EACH HAS READING HAD ITS DAY ONLINE? PLATFORM – SELECTED COUNTRIES For many years bandwidth and technical limitations meant online Sweden news was largely restricted to text and pictures. But now, in most parts of the world, it is possible to seamlessly watch news videos Spotify 40 or listen to on-demand audio as well. But what do consumers REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS SR Play 27 prefer? Reading text is convenient, but can be difficult on small Apple podcasts 17 smartphone screens, and a desire to get away from screens Slide 50 Google podcasts 8 may be one factor driving the current boom in audio listening, Website (stream) 7 according to research (Newman 2018). Castbox 4 We find that, on average, across all countries, people still prefer Podbean 4 PROPORTION THAT ACCESSED ONLINE NEWS VIDEO reading news online, but a significant proportion now say they Podcast Addict 4 IN THE LAST WEEK – ALL MARKETS prefer to watch, with around one in ten preferring to listen. Parts TuneIn 4 of the world with strong reading traditions such as Northern TUR 95 Overcast 3 Europe are most keen on text (54%), while our2020 sample of Asian KEN 93 Spreaker 3 and American markets is more equally split. The Philippines PHL 89 Audible 2 and Hong Kong are two markets where a majority say they Public broadcaster app HK 89 Pocket Casts 2 prefer to watch news online rather than read (55% and 52% MEX 88 0% 25% 50% REUTERSrespectively). INSTITUTE Across FOR markets THE STUDY we OF also JOURNALISM find that / FURTHERpeople ANALYSISwith lower 16TH 84 levels of education are more likely to want to watch online news, BRA 81 Australia Slidecompared with49 the better educated– a finding which reflects ZA traditional offline preferences around television and print. BGR 80 Spotify 33 TWN 79 Apple podcasts 21 JUNE CHL 79 ABC Listen 16 PROPORTION THAT PREFER TO READ, WATCH OR LISTEN MYS 79 Google podcasts 15 TO NEWS – SELECTED REGIONS KOR 77 Audible 10 Northern 8 54 31 7 8 CRO 77 Pandora EuropeTIME) Castbox 7 GRE 76 Southern Website (stream) 7 Europe 51 34 10 5 POL 75 6 Deezer Central and ROU 73 47 38 11 4 6 South America Pocket Casts SVK 71 RadioPublic 6 Asia 46 39 8 7 HUN 70 Overcast 5 ARG 70 Podcast Addict 5 (BRITISH North America 42 39 10 9 SGN 69 Spreaker 5 SPA 68 TuneIn 5 0% 25% 50% 75% 100% 64 Podbean 4 Read news Watch news Listen to news Don’t know CZE Public broadcaster app 63 Stitcher 3 00.01 ITA POR 62 0% 25% 50% PROPORTION THAT PREFER TO READ, WATCH OR LISTEN REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS 62 Slide 59 TO NEWS BY EDUCATION – ALL MARKETS CAN POD2. Which of the following apps or websites do you mainly use to find and play podcasts? 62 Base: All that listened to podcasts in the last month: Sweden = 741, Australia = 690. IRE High education 56 32 7 4 USA 61 AUS 60 Spotify’s move into Low education 41 41 9 9 SUI 57 podcasts, which includes 55 commissioning its own 0% 25% 50% 75% 100% NOR 54 high-quality original Read news Watch news Listen to news Don’t know FIN content, is bringing audio AUT 53 EMBARGOED: OPTQ11D_2020. In thinking about your online habits around news and current affairs which of 53 programming to a wider the following statements applies best to you? Base: Northern Europe = 22281, Southern Europe = BEL and more mainstream 8071, Central and South America = 8093, North America = 4057, Asia = 11387, high education = 30258, JPN 51 low education = 12993. audience but it is also NLD 49 raising new questions for 48 67% FRA consume video news public broadcasters. Many worry that aggregators will take much Perhaps more surprisingly, we find that in a number of countries SWE 48 online weekly of the credit/attribution for public content and are building up (average of 40 including the UK, Australia, France, and South Korea, younger GER 43 their own platforms instead. Some are withholding content from people (under-35s) are more likely to say they prefer to read rather markets) DEN 41 Spotify and Google or are previewing it first in their own apps. than watch news online. We know that the young consume more UK 39 Broadcasters have also been losing presentation and production video news than older people because they are more exposed via talent to newspaper groups, platforms, and independent studios. Facebook and YouTube, but the speed and control that comes 0% 25% 50% 75% 100% through reading often seems to trump this when it comes to Q11_VIDEO. Thinking about consuming online news video (of any kind) over the last week, underlying preference. which of the following did you do? Base: Total sample in each market ≈ 2000, Taiwan = 1027. Note: Turkey and Kenya are urban samples that will inflate figures. 100%

75% 76 28 / 29 61 50% 47

25% 23 25 VIDEO CONSUMPTION ACROSS COUNTRIES Across countries over half (52%)11 access video news via a third-party platform0% each week, such as YouTube, Facebook, and Twitter, with Looking at absolute consumption of different kinds of video news REUTERSa third (33%) INSTITUTE accessing FOR THE via STUDY news OFwebsites JOURNALISM and apps. / FURTHER But again, ANALYSIS we we also see interesting regional differences in line with these find very significant differences between markets. In Hong Kong stated preferences. Nine in ten of the online population say they Slidethree-quarters 51 (76%) access video news via third-party platforms access video news online weekly in Turkey (95%), Kenya (93%), but this figure is less than a quarter (23%) in the UK. REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS the Philippines (89%), and Hong Kong (89%), but only around Slidehalf this proportion 50 do in Northern European countries such as Germany (43%), Denmark (41%), and the UK (39%). PROPORTION THAT ACCESSED ONLINE NEWS VIDEO VIA EACH PLATFORM IN THE LAST WEEK – HONG KONG AND UK

PROPORTION THAT ACCESSED ONLINE NEWS VIDEO IN THE LAST WEEK – ALL MARKETS HONG KONG UK

TUR 95 Facebook 52% Facebook 15% KEN 93 YouTube 45% YouTube2020 8% PHL 89 Others 29% Others 7% 100% HK 89 MEX 88 75% 16TH BRA 84 76 ZA 81 50% 61 BGR 80 47 TWN 79 25% JUNE CHL 79 23 25 MYS 79 11 0% KOR 77 Consumed Consumed Consumed Consumed Consumed Consumed 77 via social news- no news- via social news- no news- CRO TIME)or other related related or other related related GRE 76 platform video video in the platform video video in the onsite last week onsite last week POL 75 ROU 73 Q11_VIDEO. Thinking about consuming online news video (of any kind) over the last week, which of the following did you do? Base: Total sample: Hong Kong = 2023, UK = 2011. SVK 71 HUN 70 ARG 70 The popularity of social networks and video platforms in Asia, Latin (BRITISH REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS SGN 69 America, and Africa seems to Slidebe stimulating 52 video consumption at SPA 68 the expense of text – even if most people consume a mix of the two. CZE 64 Even in Europe a number ITA 63 00.01 of publishers have stepped 62 POR up investment in video 62 CAN formats. The German IRE 62 public broadcaster ARD USA 61 has recently made vertical AUS 60 video a central feature of SUI 57 its Tagesschau app. NOR 55 Swiss publisher Ringier has launched Blick TV, a 15-minute news show that is broadcast online but updated through the day, while FIN 54 rival 20 Minuten is significantly increasing its video production. AUT 53 The big dilemma for publishers, however, is how to monetise video, BEL EMBARGOED:53 which has proved difficult for shorter news items. Publishers 51 JPN like the New York Times, Vox, and BuzzFeed have been focusing NLD 49 on longer form linear commissions aimed at online streaming 48 67% FRA consume video news services and cable TV. SWE 48 online weekly (average of 40 Consumer preferences around video and audio are changing, GER 43 markets) opening up new possibilities for publishers, but shifting resources DEN 41 from text carries significant risks while the commercial returns UK 39 are still far from proven. 0% 25% 50% 75% 100%

Q11_VIDEO. Thinking about consuming online news video (of any kind) over the last week, which of the following did you do? Base: Total sample in each market ≈ 2000, Taiwan = 1027. Note: Turkey and Kenya are urban samples that will inflate figures. REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS Slide 56

PROPORTION THAT USED EACH DEVICE TO ACCESS NEWS IN THE LAST WEEK (2013–20) – UK

Smartphone Computer Tablet Smart Speaker

75% 72% Smart speaker 62% Tablet

Reuters Institute for the Study of Journalism / Digital News Report 2020 50% Computer 30 43% 75% Smartphone Smart speaker 29% 25% 26% Tablet 16% 50% Computer Both Instagram and WhatsApp were designed from the ground up 3% 5% SOCIAL MEDIA PREFERENCES ALSO BECOMING 0% Smartphone MORE VISUAL for mobile usage, which gives them a cutting edge with younger 2013 2014 2015 2016 2017 2018 2019 2020 people. In developing countries, Facebook has often made 25% Q8b. Which, if any, of the following devices have you used to access news in the last week? With video now a key component of social platforms, it is interesting arrangements to bundle these applications with free data which Total UK sample 2013–20 ≈ 2000. to see how this is playing out in the networks people use most often. REUTERShas also INSTITUTEhelped them FOR THEgrow. STUDY After OF JOURNALISMyears of stagnation, / FURTHER Twitter ANALYSIS has also seen significant growth in a small number of countries, also 0% 2013 2014 2015 2016 2017 2018 2019 2020 Each year we track the importance of different social networks Slidedriven by younger 54 demographics engaging with the platform averaged across more than a dozen countries we have been more often. Twitter has also been pushing new video formats and tracking since 2014. Facebook and YouTube remain by far the most monetisation of live video in the last few years. important networks, with around two-thirds using them each week for any purpose. Almost half now use WhatsApp (48%), part of a PROPORTION THAT USED EACH DEVICE TO ACCESS NEWS shift towards more private messaging which started in around 2016. IN THE LAST WEEK (2013–20) – UK

Another big change in the last few years has been the growth Instagram for news Twitter for news Smartphone Computer Tablet Smart Speaker of Instagram which popularised visual formats like ‘stories’ and Brazil 30% (+4) Spain 20% (+4) 75% short videos via IGTV. Instagram now reaches more than a third 72% Smart speaker Chile 28% (+8) USA 202017% (+2) 62% (36%)REUTERS weekly INSTITUTE and FORtwo-thirds THE STUDY of under-25s OF JOURNALISM (64%). / FURTHER As people ANALYSIS spend Tablet Italy 17% (+4) Brazil 17% (+2) more time with the network, the role of news has also increased 50% Computer Spain 17% (+5) Japan 15% (+5) 43% Slidesignificantly. 53 Instagram reaches 11% across age groups, almost as Smartphone many as use Twitter for news. 29% For more on the top social networks see16TH section 3: Country and Market 25% 26% pages 16% 3% 5% PROPORTION THAT USED EACH SOCIAL NETWORK FOR By allowing Instagram and WhatsApp to develop separately, 0% ANY PURPOSE IN THE LAST WEEK (2014–20) – AVERAGE OF 2013 2014 2015 2016 2017 2018 2019 2020 Facebook as a company has been able to cater for many different 12 COUNTRIES demographics and tryJUNE new formats, without losing its core loyalists. Q8b. Which, if any, of the following devices have you used to access news in the last week? Facebook Facebook Messenger Twitter Instagram Across countries, Facebook Inc now reaches 85% each week on Total UK sample 2013–20 ≈ 2000. Snapchat WhatsApp YouTube average, rising to 94% in Brazil and 96% in Kenya and South Africa. 75% YouTube 63% 60% TIME) 27% 60% 61% DEPENDENCEWhatsApp ON SMARTPHONES CONTINUES 53% 24% 48% TO GROWSnapchat 45% 38% 36% OverInstagram two-thirds (69%) of people now use the smartphone for 30% 31% news weekly and, as we’ve seen, these devices are encouraging 23% Twitter 19% the growth of shorter video content via third-party platforms as 15% 17% (BRITISH13% wellFacebook as audio content Messenger like podcasts. Those who use smartphones 8% as a main device for news are significantly more likely to access 5% Facebook 0% news via social networks. 2014 2015 2016 2017 2018 2019 2020 Usage is often highest in parts of the Global South such as Kenya Q12a. Which, if any, of the following have you used in the last week00.01 for any purpose? Base: Total 2014-19 sample across 12 countries ≈ 24000 (~19000 in 2014). Note: From 2015-20 the 12 countries (83%) and South Africa (82%) where fixed-line internet tends to included are UK, USA, Germany, France, Spain, Italy, Ireland, Denmark, Finland, Japan, Australia and Brazil. In 2014, we did not poll in Australia or Ireland. REUTERSbe less prevalent. INSTITUTE AccessFOR THE is STUDY lowest OF in JOURNALISM Canada (55%), / FURTHER Japan ANALYSIS (52%), REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS Slideand in much 55 of Eastern Europe, though even here the smartphone Slide 57 PROPORTION THAT USED EACH SOCIAL NETWORK FOR NEWS has become – or is on its way to becoming – the main platform for IN THE LAST WEEK (2014-20) – AVERAGE OF 12 COUNTRIES accessing news.

Facebook Facebook Messenger Twitter Instagram Snapchat WhatsApp YouTube PROPORTION THAT USED A SMARTPHONE TO ACCESS NEWS PROPORTION THAT USE SMART SPEAKERS FOR ANY PURPOSE IN THE LAST WEEK (2013-20) – SELECTED COUNTRIES (2017-20) – SELECTED COUNTRIES 40% 36% 36% YouTube USA UK Germany France Japan Spain USA Germany Canada Australia South Korea UK EMBARGOED: 75% WhatsApp 30% 73% 20% Spain 19% UK Snapchat 62% 59% Japan South Korea 21% 58% 58% 13% 13% 20% Instagram 52% 12% 12% Australia 16% 50% France 16% 10% 9% Canada 12% Twitter35% Germany Germany 11% 30% 10% 9% USA 7% Facebook29% Messenger 6% 8% 25% 25% UK 22% 3% 0% 8% 1% Facebook19% USA 2017 2018 2019 2020 0% 2014 2015 2016 2017 2018 2019 2020 ALSO: 0% Spain 9%, France 4%, Sweden 4%, Norway 4%, Mexico 4%, Japan 4% 2013 2014 2015 2016 2017 2018 2019 2020 Q12b. Which, if any, of the following have you used in the last week for news? Base: Total 2014–19 sample across 12 countries ≈ 24000 (~19000 in 2014). Note: From 2015–20 the 12 countries included are UK, USA, Germany, France, Spain, Italy, Ireland, Denmark, Finland, Japan, Australia and Brazil. In 2014, Q8b. Which, if any, of the following devices have you used to access news in the last week? Q8a. Which, if any, of the following devices do you ever use (for any purpose)? Base: Total we did not poll in Australia or Ireland. Base: Total sample in each country 2013-20 ≈ 2000. sample in selected markets 2017–20 ≈ 2000. REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS Slide 56

PROPORTION THAT USED EACH DEVICE TO ACCESS NEWS IN THE LAST WEEK (2013–20) – UK

Smartphone Computer Tablet Smart Speaker

75% 72% Smart speaker 62% Tablet

50% Computer 30 / 31 43% 75% Smartphone Smart speaker 29% 25% 26% Tablet 16% 50% Computer Across countries almost half (48%)3% use two or more devices5% to 0% CONCLUSIONSmartphone access news2013 each2014 week 2015compared2016 with2017 39%2018 in 2014.2019 Computers2020 25%and laptops remain important for many but the convenience The COVID-19 lockdown has reminded us both of the value of Q8b. Which, if any, of the following devices have you used to access news in the last week? Totaland UK versatility sample 2013–20 of ≈the 2000. smartphone continues to win out. In the UK media that bring us together, as well as the power of digital the smartphone overtook the computer in 2017 and is now used networks that connect us to those we know and love personally. by0% around two-thirds of our sample. Tablets are flat in terms of For many of those stuck at home, television remains a window to 2013 2014 2015 2016 2017 2018 2019 2020 usage for news (26%) with a small group of older and richer users the wider world, and some broadcasters have provided a platform continuing to value their larger screens. Smart speakers are now for governments and health authorities to communicate health used for news by around 5% of online users. and other advice to mass audiences.

New digital behaviours have also emerged in this crisis that are PROPORTION THAT USED EACH DEVICE TO ACCESS NEWS likely to have long-term implications. Many have joined Facebook IN THE LAST WEEK (2013–20) – UK or WhatsApp groups for the first time and have engaged in local Smartphone Computer Tablet Smart Speaker groups. Young people have consumed more news through services like Instagram, Snapchat, and TikTok. Video conferencing has 75% Smart speaker 72% emerged as a new platform for personal communication2020 but has 62% Tablet also changed the face of government press conferences. The media 50% have embracedComputer these new technologies in terms of remote working, 43% but also Smartphonein terms of the production and distribution of content. 29% 25% 26% 16TH 16% The biggest impact of the virus is likely to be economic, with local and national media already cutting staff or publishing less 3% 5% 0% 2013 2014 2015 2016 2017 2018 2019 2020 frequently. The coronavirus crisis is driving a cyclical downturn in the economy hurting every publisher, especially those based Q8b. Which, if any, of the following devices have you used to access news in the last week? on advertising, and likelyJUNE to further accelerate existing structural Total UK sample 2013–20 ≈ 2000. changes to a more digital media environment in terms of audience behaviour, advertising spending, and reader revenues. Reader payment alternatives such as subscription, membership, and AI-driven smart speakers such as the Amazon Echo and Google donations will move centre stage, but as our research shows, this Home continue to become more widely available, along with the TIME) is likely to benefit a relatively small number of highly trusted intelligent assistants inside them. They are now in more than 20 national titles as well as smaller niche and partisan media brands. countries including India, Mexico, and a wider range of European The crisis in local media will become more acute with more calls languages. Usage for any purpose is up five percentage points to 19% for support from government and technology companies – with in the UK and a similar amount in Germany to 12%. In South Korea, all the problems that this entails in terms of media independence. where local firms make a range of popular speakers including the Naver Wave and Kakao Mini, usage is up four points to 13%.(BRITISHAs if this were not enough, dependence on aggregated and mobile content has made it harder for news brands to forge direct Both Amazon and Google have been looking to improve the news relationships with consumers. Our report shows that younger experience after disappointing levels of take-up for on-demand users, especially those now coming into adulthood, are even less briefings. Google has been rolling out a new atomised audio news connected with news brands and more dependent on social media. service that aims to offer a more story-based 00.01approach and the BBC More effective distribution of formats like video, podcasts, emails, has added more controls to its news briefing allowing consumers REUTERS INSTITUTE FOR THE STUDY OF JOURNALISM / FURTHER ANALYSIS and notifications may help but most publishers continue to struggle to skip between stories. Despite these initiatives, our data suggest to engage more deeply with the next generation and other hard-to- that only between a third and a quarter are using their devices to Slide 57 reach groups. More widely, concern about misinformation remains regularly access news. high, while trust in the news continues to fall in many countries.

Despite this, there are some signs of hope. The COVID-19 crisis has PROPORTION THAT USE SMART SPEAKERS FOR ANY PURPOSE clearly demonstrated the value of reliable trusted news to the (2017-20) – SELECTED COUNTRIES public but also to policymakers, technology companies, and others USA Germany Canada Australia South Korea UK who could potentially act to support independent news media. The EMBARGOED: creativity of journalists has also come to the fore in finding flexible 20% 19% ways to produceUK the news under extremely difficult circumstances. Fact-checkingSouth Korea has become even more central to newsroom 13% 13% 12% 12% operations,Australia boosting digital literacy more widely and helping to 10% 9% counter theCanada many conspiracy theories swirling on social media and elsewhere.Germany Publishers have also found innovative ways to display and interrogateUSA data, just one of many format innovations that 0% have helped audiences understand the background and the 2017 2018 2019 2020 implications for each individual. ALSO: Spain 9%, France 4%, Sweden 4%, Norway 4%, Mexico 4%, Japan 4% The next 12 months will be critical in shaping the future of the news industry. Many news organisations go into this period clearer than Q8a. Which, if any, of the following devices do you ever use (for any purpose)? Base: Total ever about the value of their product even if the immediate outlook sample in selected markets 2017–20 ≈ 2000. looks uncertain.