In this piece
Overcoming indifference: what attitudes towards news tell us about building trust
In this pieceExecutive summary and key findings↑ | The impact of the digital media environment on trust | Background | Summary of key findings | 1. Identifying gaps in trust in news↑ | 2. Who are the ‘generally untrusting’?↑ | 3. Indifference as an underappreciated driver of low trust↑ | 4. Many are uncertain about the practices of journalism↑ | Conclusion ↑ | Implications of our findings | Focus of future research | Appendix: Methodology ↑ | References ↑ | Footnotes | About the Authors ↑ | Acknowledgements ↑
Executive summary and key findings ↑
In recent decades, trust in news has declined in many parts of the world (Fletcher 2020). While the coronavirus crisis has reminded some of the value of independent journalism, boosting trust in some places (Newman et al. 2021), many continue to regard news with considerable scepticism. The media are at the centre of often intense public arguments over how societies generally – and news specifically – deal with important and sometimes polarising issues including the pandemic but also more broadly the climate emergency, populist politicians, racial injustice, other social inequalities, and much more. One prominent feature of these debates is often outright hostile attacks on news media and individual journalists by vocal and visible critics who actively express their distrust and disdain for the media and its many shortcomings, both real and perceived, especially on social media.
To be fair, news media are not alone in facing often dwindling public trust. Trust in many other institutions, including both national and local governments, has also declined in some cases, as has interpersonal trust. However, social scientists have long stressed that, despite frequent and sweeping claims of a ‘crisis of trust’, there is no evidence for a consistent, across-the-board decline in public trust in every country, every institution, or every news organisation (Norris 2011, Newman et al. 2021). These wider developments and pronounced country-to-country differences are important for trust in news, too, because attitudes towards news media are difficult to disentangle from other forms of trust towards other institutions (Hanitzsch et al. 2018).
Trust also matters for democracy. When the public place their trust in those who are in fact trustworthy, it can be profoundly enabling. But its absence can be equally debilitating, and, when trust is misplaced, it can lead us astray. Trust in news specifically matters for journalists who want people to rely on their reporting, for news media who depend on people paying attention to (and paying for) the news they produce, and for each of us as citizens. We all need trustworthy sources of information to understand and navigate our worlds and consider perspectives outside of our own narrow personal experiences.
The impact of the digital media environment on trust
Understanding trust in news and how news media may be able to build trust is especially important in an increasingly digital, mobile, and platform-dominated media environment where more and more people rely on intermediaries, including search engines, social media, and messaging applications, to access and discover news. As more people spend more of their time using platforms – which often provide limited context on the sources of information displayed and where many do not recall the brands behind stories they have read (Kalogeropoulos et al. 2019) – there are considerable concerns about how such changing audience behaviours will impact attitudes towards news outlets that depend on trusting relationships with audiences.
The changing context around how people access and use news and information risks imparting trust where it is not deserved (Gursky and Woolley 2021), enabling the widespread dissemination of problematic information. The trust gap between news in general and news accessed via platforms also risks diminishing trustworthy brands’ standing ‘by association’, as people come across news in environments that also offer many other kinds of information, including sometimes misinformation or outright disinformation. Surveys document that majorities in many countries are concerned about whether online news is real or fake. Large numbers of people are worried about false or misleading information disseminated via Facebook or messaging applications such as WhatsApp, but also, to a lesser extent, via Google, YouTube, or less widely used social media such as Twitter (Newman et al. 2021).
In this report, part of a larger RISJ project focused on trust, we use original survey data from four countries – Brazil, India, the UK, and the US – to develop a more detailed understanding of how different segments of the public hold varying degrees of trust in news. We do so in order to help those interested in building trust in news better understand the people they are trying to reach.
We examine three groups we call the ‘generally untrusting’, the ‘selectively trusting’, and the ‘generally trusting’, defined on the basis of the relative number of news brands respondents say they trust ‘somewhat’ or ‘completely’. We find consistent gaps between groups not only in their attitudes towards news itself, but also in their views towards other institutions in society more broadly. Perhaps most importantly, across all four countries covered, we find that those who generally lack trust in news are not necessarily the most vocal and angry about news coverage (who are, on closer examination, often people who are selectively trusting towards certain news providers). Instead, the generally untrusting tend to be the least knowledgeable about journalism, the most disengaged from how it is practised, and the least interested in the editorial decisions and choices publishers and editors make daily when producing the news.
The primary challenge news media and journalists face from this part of the public is not hostility, but indifference. Reaching them, demonstrating the value journalism can hold for them, and earning their trust will call for a different set of responses than those required for engaging with more vocal and visible critics or for incrementally increasing trust among the already trusting parts of the public.
The RISJ’s larger Trust in News Project seeks to understand the drivers of trust in news, the factors responsible for its apparent decline in many countries in recent years, the differences in how this plays out in different places around the world, and what might be done about it. This report builds on two previous reports we have published: one based on interviews we conducted with senior managers and journalists at news organisations worldwide (Toff et al. 2020) and one based on qualitative conversations we held with cross-sections of news audiences (Toff et al. 2021), which focused on how media users define trust and think about the news they engage with. While we found some overlap between both practitioners and audiences in terms of the concerns they expressed about the contemporary digital news media landscape, especially on social media, we also found some key differences. Audiences were far less versed in the professional practices that underlie differences between news brands and more likely to base assessments about trustworthiness on familiarity and impressions about brand reputation. Some revealed in focus groups and interviews that they sometimes relied on cues tied to how news was presented – including its visual qualities, use of language, and other easily observable indicators of difference – as shortcuts for assessing whether a source was deserving of their trust.
These findings informed our approach in this report, as did years of social science research stressing how trust in news is tied to other forms of institutional and interpersonal trust. We designed an original survey questionnaire and fielded it in May and June 2021 in four countries spanning the Global North (the UK and the US) and the Global South (Brazil and India). While each of these large, diverse countries contains its own unique political, cultural, and societal divides, they are all grappling with the role played by digital media platforms in how their citizens stay informed.
We worked closely with three independent survey firms – Datafolha in Brazil, Internet Research Bureau in India, and Kantar in the US and the UK – to poll samples of approximately 2,000 respondents per country to reach populations broadly representative of each country’s overall population. Surveys were conducted online in all but Brazil, where respondents were interviewed by telephone and randomly selected on the basis of their cell phone numbers. Elsewhere, quotas were applied to reach population targets. In India, supplementary recruitment was conducted via telephone and the messaging app WhatsApp to reach greater numbers of non-English speakers in smaller towns and cities. Additional information about the methodology used in these surveys is provided in an appendix.
Surveys were designed to capture a mix of attitudes about journalism as it is practised in each of the four countries, along with underlying characteristics about individual respondents, including their demographics, their political views, how knowledgeable they are about news, and what they think about journalism’s role in society. We sought to balance specificity with breadth by combining some questions about individual news brands and platforms with other questions about news more generally. Survey questionnaires took approximately 12–15 minutes online; in Brazil, several questions were omitted because the questionnaire took longer to complete over the phone.
Summary of key findings
This report contains a range of findings about news audiences in each of the four countries, focusing on audiences overall as well as different segments of the public categorised according to their degree of trust towards news brands in their country. We summarise several of the key results of our analysis here:
- People are more trusting of news they themselves use, including on social media, but less trusting of news they don’t use, especially news found on digital platforms. Relatively high percentages overall say they ‘somewhat’ trust information reported in the news media generally; however, levels of trust are much lower for specific news brands and news found on social media. That trust gap between online and offline news, however, is largely driven by low trust among people who do not use news on these platforms.
- Many hold highly negative views about basic journalistic practices. Large minorities in all four countries have very negative or cynical views about how they think journalists do their jobs, including allowing personal opinions to influence coverage, accepting undisclosed payment from sources, or deliberately seeking to manipulate the public. Remarkably, these views vary only somewhat between those who otherwise exhibit low and high trust in news. Even the generally trusting often have what journalists may regard as a pretty dim view of basic journalistic practices.
- The least trusting towards news tend to be older, less educated, less interested in politics, and less connected to urban centres. Although there were some differences by country with respect to which groups tended to be more or less trusting towards news brands generally, we also find key similarities across countries in which groups typically lack trust towards individual news brands. In all four countries, we find wide disparities between the generally trusting and untrusting in whether they think news organisations understand people like them or genuinely want to hear from the public.
- The least trusting pay less attention to and are more indifferent towards specific characteristics about how journalism is practised. We find that factors involving editorial practices, including transparency about how news is produced and who reports it, were deemed less important to people who were generally untrusting towards news. People who were most trusting towards news were much more likely to say these characteristics were important when they made decisions about what news sources to use.
- Experience interacting with journalists is rare and familiarity with basic concepts concerning how news works is often low. Most said they had not interacted with journalists and few said they were especially familiar with basic terms and concepts in journalism, including the difference between an editorial and a news story, or between a news story and a press release. The least trusting were also the least likely to have experience contacting or conversing with journalists or knowledge about journalistic terminology.
- Gaps in trust in news align with deficits in social and interpersonal trust as well as dissatisfaction with democracy. We find a strong correlation between levels of trust in news, the degree to which people are bonded to other individuals and specific groups in society, and how satisfied people are about the way democracy is working. These results point to the extent to which trusting attitudes towards news are in part driven by factors external to the news itself. Our data also show that low trust in news may have implications for how people think about the media environment more widely: the generally untrusting are somewhat less supportive of free expression and somewhat more in favour of government censorship compared to other respondents.
1. Identifying gaps in trust in news ↑
In this section, we begin by quantifying how much people trust the news media overall in Brazil, India, the UK, and the US, the four countries that are the focus of our project. First, we discuss some of the challenges associated with measuring trust and detail results for different sources of news in each country. Next, we use these data to divide audiences in each country into three broad groups: those who are ‘generally trusting’ of most or even all brands in their country; those who are ‘generally untrusting’ of most or even all brands; and those in between, whom we call the ‘selectively trusting’. In the last part of this section – and elsewhere in this report – we use this categorisation scheme to further investigate factors related to apparent gaps in trust among segments of the public. Some of these gaps are reflective of concerns about specific news sources and journalistic practices in each of the four countries, but others appear rooted in people’s perceptions about the information environment more generally – namely, the role played by social media platforms as well as more generalised distrust towards a host of other institutions in society. Attitudes about news specifically are not easily disentangled from these broader considerations.
As we have previously written (Toff et al. 2020; Toff et al. 2021), trust in news is a multifaceted concept involving not only people’s subjective attitudes about how journalism is practised, but also differing expectations around how news should interact with other institutions in civic life. Trust is, by definition, a relationship (see Schiffrin 2019; Tsfati and Cappella 2003), and, as such, the concept often means something different to different people, which makes it particularly difficult to measure consistently in surveys.
For that reason, social scientists who study trust in news have often taken a range of different approaches, which have included asking specifically about different dimensions of trust (e.g. whether news is accurate or fair) as well as different objects of trust (e.g. journalists, brands, news on specific platforms, or news in general). (See Figure 1)
In designing our own survey, we sought to balance both of these considerations. We include questions focused on different objects of trust – news in general as well as specific news brands – and we also incorporate questions allowing us to investigate the relationship between aggregate levels of trust and how such attitudes align with other identities, ideologies, and experiences.
In practice, survey self-reports about different dimensions of trust in news tend to be highly correlated with one another – that is, people are not always very good at separating out how they feel about a particular source of news independently from how they might evaluate aspects of its journalistic performance. For that reason, although we ask many other questions designed to capture what people think about news and the social and political forces around them, we begin with some basic questions that explicitly ask people to reflect on trust – how much they trust news in general, news they personally choose to use, and then specifically how much they trust a list of prominent news brands and digital platforms widely used in their country. Elsewhere in the report, we will say more about the different dimensions of trust that we think underlie these assessments.
Figure 1: A framework for measuring trust in news
Note: The categorisation of objects of trust is drawn from Strömbäck et al. (2020). In our previous report (Toff et al. 2021), based on qualitative interviews with individuals in each of the four countries, we argued that impressions, preconceptions, gut feelings, and internalised messages about news often played an underappreciated role in how people formed trusting or distrusting attitudes towards specific media.
Most people say they ‘somewhat’ trust information in the news
While previous surveys, including the RISJ’s Digital News Report (Newman et al. 2021), show majorities in most countries reject the notion that you can trust ‘most news most of the time’, we find this broad scepticism towards news does not necessarily prevent people, on balance, from concluding that they still have some trust in what gets reported.
We find large majorities in all four countries say they at least ‘somewhat’ trust information in the news media. While very few say they are ‘completely’ trusting, except in India where 38% say so, the percentages who say they do not trust information in the news media is considerably smaller than might be expected based on past research.
Our results likely differ from other studies because of the specific way we approached asking about trust. First, our questions ask people to state how much they trust ‘information from the news media’ rather than news media as institutions. The emphasis on ‘information’ in these questions is intentional. As Strömbäck and colleagues (2020) argue, unlike other aspects of trust, how the public evaluates information reported by the news media matters in ways that are fundamental for democracy. Whether people see news media as a reliable source of information specifically determines whether they see the press as ‘living up to its ideals’ when it comes to both its role in democratic life and serving communities more generally.
We asked respondents about both whether they trust information from the news media in their country in general and the news media they themselves choose to use. We further included specific brand-level evaluations of trust in the information reported by 15 news brands in each country, selected for being widely used by a variety of different audience segments.1 Finally, in a separate battery of questions, respondents were asked about their trust in news found on major digital media platforms such as Facebook, WhatsApp, Twitter, and YouTube, since such platforms play an increasingly important role in how people engage with news (Strauß et al. 2020).
The second way our approach to measuring trust differs from others is that we do not include a middle response category when asking about trust. In other words, when respondents report how much they trust information from a variety of sources, they use a four-point scale ranging from ‘trust completely’ and ‘trust somewhat’ to ‘do not trust very much’ and ‘do not trust at all’. This style of questioning – a forced-response structure – prevents people who may feel somewhat ambivalent or unsure from staying on the fence and selecting a midpoint like ‘neither trust nor do not trust’ when they actually do hold an opinion one way or the other.2
The relatively high levels of modest trust we find using our approach are meaningful: people in general are not rejecting everything they see in the news as unreliable. However, saying you are ‘somewhat’ trusting towards information being reported in the news media is a low bar, and elsewhere in the report we show how even those who are generally trusting towards most news brands hold negative views about how journalists do their work. Instead, these findings underscore the degree to which people hold a mix of trusting and distrusting views.
Trust in specific brands and news on platforms is lower than trust in news in the abstract
When we examine brand-level trust alongside overall trust in the abstract, we find three interesting patterns. First, there is considerable variation in brand-level trust in each country, with several brands in most countries viewed as untrustworthy. Again, the one exception is India, where a plurality say they trust all 15 brands somewhat or completely.
Second, trust in news in the abstract is generally higher than for any specific news brand. Even the most trusted brands, like the BBC in the UK, receive slightly lower percentages saying they trust it somewhat or completely (75%) compared to news overall in the country (78%). We think this is a consequence of specific brands often being more polarising than information in the abstract; when asked about news in general, people may default to thinking about more established, respected sources of information rather than averaging across the entirety of the news media landscape or thinking about exceptions where they might have more cause for scepticism (see Daniller et al. 2017).
Third, we also find that trust in news on most digital platforms is lower than trust in specific news brands. While news on Facebook is trusted ‘somewhat’ or ‘completely’ by a majority in India (65%), rates are considerably lower in Brazil (42%), the US (35%), and the UK (29%). Gaps between trust in specific news brands and trust in news found on platforms is particularly wide for WhatsApp as well, where news is trusted somewhat or completely by just 28% in the UK and 32% in the US, but by 57% in India and 45% in Brazil. Google, on the other hand, ranks highest, as the most trusted platform for news in all four countries. In Brazil and India the percentage who say they trust news on Google exceeds any single brand in either country.
1.2 Changing ways people encounter news and the role of partisanship
One challenge with interpreting these results from one country to the next is that the news media in each country are distinct from one another in several ways. The specific forms and modes of media people use both online and offline differ considerably. Likewise, differences in the partisan political context in each country also shape attitudes towards news in specific ways.
People are generally more trusting of sources of news they themselves use
Certain forms of news are used more widely in some countries compared to others. In India, for example, 84% said they got news sometimes or often from printed newspapers or magazines, whereas in Brazil just 22% said they did the same. While social media was used widely in all four countries as a source of news, messaging apps were far more likely to be used as a source of news in Brazil (71%) and India (82%), but far less so in the UK (32%) and US (30%).3
Depending on which forms of news people used, it tended to affect how much people said they trusted news found through that specific intermediary.4 Those who used social media platforms, and especially those who said they got news from them, generally rated the news found on those digital platforms as more trustworthy compared to those who did not use the platforms.
On the one hand, these differences are not surprising since people may steer clear of using platforms they do not trust. On the other, some gaps in trust in news on platforms may be driven by impressions and assumptions about news found there based on elite cues or word-of-mouth about their reputations, rather than actual experiences using these platforms.
Importance of partisanship for specific brands
In addition to differences in media mode, one of the other major factors that drives differences in brand-level trust is partisanship. While considerable previous research has focused on party cueing as a factor contributing to gaps in trust in news (e.g. Ladd 2011), much of this work has focused only on the US, where partisan polarisation is a well-documented and debated phenomenon.
We, too, find considerable partisan differences in levels of trust towards brands in the US for all but one of the 15 brands we asked about. All exhibited a more than double-digit percentage point gap between the proportion of Democrats and Republicans who said they trusted information from the brand somewhat or completely.5 Among the 15 brands, the most polarising was CNN, where there was a 50 percentage point gap between party identifiers: 79% of Democrats said they trusted CNN somewhat or completely compared to 29% of Republicans.6 Similarly large gaps were apparent with respect to the New York Times (trusted by 78% of Democrats but 33% of Republicans) and Fox News (36% versus 62%).
We find evidence of similar trends in the UK, India, and Brazil; however, partisan divides in these countries were often found only for specific brands rather than across the board. The BBC, for example, was trusted somewhat or completely in the UK by a roughly equal percentage of Labour and Conservative party supporters – 78% and 79%, respectively. Popular titles like the Daily Mail exhibited more polarising responses; it was trusted somewhat or completely by just 41% of Labour supporters compared to 63% of Conservatives. In India, the most polarising brands were Republic TV/Republic Bharat (trusted by 72% of Bharatiya Janata Party (BJP) supporters but by only 50% of all others) and Zee News (trusted by 85% of BJP supporters compared to 63% of non-supporters). On the other hand, trust in NDTV was relatively high across the board, trusted somewhat or completely by 75% of BJP supporters and by 81% of all others.
In Brazil, where fewer people historically identify with political parties (65% of our sample did not do so), we nonetheless found a significant gap along partisan lines in attitudes towards Globo, the conglomerate behind TV Globo, Globo News, and G1, among other brands. An overwhelming majority (83%) of supporters of the Worker’s Party (PT) in Brazil, also called Petistas, said they trusted information reported by the brand somewhat or completely compared to only a slim majority (54%) of all others. This is a shift from what happened when the party held the presidency. At that time, PT supporters tended to be very critical of Brazilian media in general, especially Globo.7
We return to this subject of partisanship and political attitudes in sections 2 and 4 of this report.
1.3 Dividing up respondents by levels of trust
If many people hold a mix of attitudes towards different news brands, what does it mean then to be highly trusting or distrusting of news in general? Much of this report focuses on the factors that are associated with different levels of trust, but classifying respondents by trust is not as straightforward as it might seem.
Distinguishing the ‘generally untrusting’ from the ‘selectively’ and ‘generally trusting’
As we argued in our previous report (Toff et al. 2021), one of the reasons why trust is important for individual citizens is because it helps them navigate the information environment. Not every news source is deserving of trust or consistently reports factually accurate information reliably. A well-informed, thoughtful news consumer is discerning about the sources they encounter; they know which sources of news they can trust but they also know which to be sceptical of and why. In other words, blanket trust of all news may be just as unhealthy as generalised distrust of all sources.
With that in mind, we sought to classify individuals in each country using the number of news brands respondents said they trusted somewhat or completely. Doing so allows us to differentiate between people who say they are trusting (or not trusting) across the board almost regardless of the brand.
In all four countries, we employ a categorisation scheme that segments respondents into one of three groups, which we call the ‘generally untrusting’ (who trust information from a below-average number of brands in their country), the ‘generally trusting’ (who trust information from an above-average number of brands), and the ‘selectively trusting’ (those in between, who trust some but not necessarily all or most of the brands asked about in their country).
We use the term ‘untrusting’ as opposed to ‘distrusting’ deliberately to denote an absence of trust, which may be driven not only by affirmative distrust of specific brands but also a more general lack of brand awareness. In other words, respondents who have never heard of most or even all of the 15 news brands are classified as ‘generally untrusting’ along with those who specifically report not trusting any of them. What we seek to capture using this method are distinctions between those who have a trusting relationship with an average or above-average number of news brands in their country from those who do not.
This method of categorising respondents allows us to make more standardised comparisons between countries, and we do so throughout the remainder of this report as we investigate characteristics associated with people in each of these groups. We acknowledge, however, that this method also oversimplifies some of the country-specific differences at the extremes. For example, in India, even among the ‘generally untrusting’ there were very few respondents who said they did not trust any of the 15 brands, as well as a majority of the ‘generally trusting’ group who said they somewhat or completely trusted each and every one of them. In the US, on the other hand, 12%, or nearly half of the ‘generally untrusting’, trusted none of the 15 brands versus just 7% who trusted all of them. We do not believe the boundaries between groups are black and white, but we do believe segmenting respondents into these three groups is a useful mechanism for characterising relative differences within these populations that are revealing in important ways.
People who lack trust in news also lack trust towards other institutions
Not all gaps in trust in news can be explained by attitudes towards news itself, which is the focus of much of the remainder of this report. In fact, when we divide up respondents according to the ‘generally untrusting’, ‘selectively trusting’, and ‘generally trusting’, we find that these divides also map onto attitudes towards other institutions in society.
First, we present results for each country overall. We included in our survey a set of questions designed to gauge levels of trust towards various groups in each country, adapting a commonly used battery of questions in the long-running World Values Survey (WVS) that asks respondents how confident they are in various institutions. The specific groups or organisations we included in our survey differed to some degree depending on the national context, but we included many across all four, including the military, the courts, local or national governments, and scientists, as well as the press. (We use the term ‘the press’ to remain consistent with the way it is referred to in the WVS.) We find in most countries that the press was among the least-trusted institutions named, but we also find that majorities in all but the UK said they trust the press at least somewhat.
Trust in the press, however, is far from uniform. When we examine responses of those who are generally untrusting of news brands versus those who are generally trusting, we find wide disparities. In the US, for example, as few as 10% of those who are generally untrusting say they ‘completely’ or ‘somewhat’ trust the press, compared to 85% of those who are generally trusting of most news brands. Gaps in trust towards the press between the generally untrusting and trusting were somewhat smaller in the UK (by 56 percentage points) and Brazil (45 percentage points), and smallest overall in India (36 percentage points).
What is also striking about gaps between the generally untrusting and trusting is that respondents in each of these groups also hold divergent attitudes towards other institutions in society beyond just the press. For example, while 69% of Americans say they trust the courts somewhat or completely, just 50% of those who are generally untrusting of news brands say the same, versus 84% of those who are generally trusting of news.
Similar divides are seen across most other institutions in the other three countries as well. When we average across these groups, the gap in trust between the generally untrusting and generally trusting (of news) is smaller than the gap in trust towards the press, but it is still more than 30 percentage points in the UK and US, albeit smaller in Brazil (11 percentage points) and India (23 percentage points).8
What these results suggest is that some divides within countries around trust in news may be driven, at least in part, by more fundamental differences in worldviews when it comes to a broader set of institutions in society. In sections 2, 3, and 4, we delve more deeply into some of these differences. They involve not only assessments about how journalists conduct themselves, but also differing attitudes about what role news organisations should play in public life.
2. Who are the ‘generally untrusting’? ↑
In this section, we explore some of the key divides across demographic, socioeconomic, and political lines when segmenting respondents by levels of trust in news, in order to better understand who tends to be the least trusting towards news brands overall. We find some patterns that vary from country to country involving people’s social and political identities, but we find some more consistent patterns as well: those who trust news the least also tend to be the most dissatisfied with democracy. We also find a link between a lack of interest in politics and a lack of trust in news, which suggests that part of what may explain variation in levels of trust in news is indifference towards news rather than active hostility – a point we return to in the next section. Shedding light on these divides illuminates how broader structural factors may contribute to gaps in trust more generally.
2.1 Demographic, political, and place-based trust gaps
In examining differences between the generally untrusting and other groups, we find overall that each country is rather unique, with more differences than similarities across particular subgroups from one country to the next. This is especially true when it comes to the role of politics, where each country’s distinct environment in terms of press/state relations shapes attitudes towards news in different ways. Gaps in trust can clearly be traced to attitudes people hold towards polarising political leaders or parties, but not in a uniform ideological manner. We also find some more consistent trends with respect to particular demographic variables: older people, those without college degrees, and, to a somewhat less consistent extent, people who are white or living in smaller towns or rural areas all tend to be more concentrated among the group we call the ‘generally untrusting’.
Identifying demographic and socioeconomic divides in trust in news
The accompanying table summarises key differences when comparing the generally untrusting to other groups in each of the four countries. We find the most consistent patterns with respect to age: people aged 55 and older are more often generally untrusting, whereas those who are under 35 are typically overrepresented among the generally trusting segments of the public.9 These differences are especially notable in Brazil, where people over 55 make up 38% of the generally untrusting group but only 18% of those who are generally trusting. The reverse is also true: people under 35 account for 42% of the generally trusting group in Brazil but just 24% of the generally untrusting.
In terms of gender, we find consistent patterns across Brazil, India, and, to a lesser extent, the UK, where the generally untrusting are more likely to be men (accounting for 59% of the group in both Brazil and India and 53% in the UK). In the US, the generally untrusting are about evenly divided by gender, although men are significantly more likely to say they trust all 15 brands.
When it comes to socioeconomic status (as measured by education, employment, and household income) we see some broadly similar trends across all four countries. In India, the UK, and the US, for example, people without college degrees are more often generally untrusting and less likely to be classified as generally trusting; however, this pattern did not extend to Brazil, where differences were not significant.10 We also find in all four countries that individuals who are self-employed tend to skew more towards being generally untrusting. Other groups that tend to be more untrusting include those who are retired or unemployed in India as well as those who are not private sector workers in the US and the UK. For household income, we find that the generally untrusting are less likely to be found among those with the lowest household incomes in Brazil but somewhat more likely to be found among those in the UK and India.11 At the other end of the spectrum, in India, the UK, and the US, those in the highest income brackets are also more likely to be found among the generally trusting.12
We find even more differences from one country to the next when we consider religion, race, and ethnicity (or caste in the case of India) even though differences between smaller racial and religious subgroups can be difficult to detect due to sampling limitations. In Brazil and the UK, we find that people who are not affiliated with any religion tend to be more generally untrusting.13 In India, Hindu respondents tend to be underrepresented among the untrusting compared to non-Hindu respondents, and in the US, Protestants and Evangelical Christians were overrepresented among the generally untrusting. With respect to race, people who identify as white are more likely to be generally untrusting in Brazil, whereas those who identify as Black, brown, or indigenous in Brazil tend to be overrepresented among the generally trusting.14 Similarly in the UK, people who identify as Black, Asian, and minority ethnic are more likely to be generally trusting rather than untrusting; in the US, no significant differences were found between trust groups along racial or ethnic lines. In India, those who identify with the General/Upper caste category are less likely to be generally untrusting.
Political drivers of differences in trust in news
Some of the demographic and socioeconomic differences we describe above map onto political divides in each of the four countries, so it is no surprise then that we find some stark internal divides in trust across political lines, visible in both partisanship and how people evaluate country leaders.
Partisanship is an especially divisive characteristic in the US. When we compare the percentage of the overall sample that identify with or lean towards the Republican Party (33%) versus the Democratic Party (44%), we find the generally untrusting as a group skew far more Republican (54%) than Democrat (16%). These differences are seen in the reverse proportions among the generally trusting, where Republicans account for just 22%, versus Democrats who are 61% of the generally trusting.
Partisan differences are also visible to a lesser extent in Brazil and India but only minimally so in the UK. In India and Brazil, two countries with large numbers of political parties and many voters who said they do not identify with any party at all, we find that supporters of the BJP, which leads the current right-leaning government in India, and the Worker’s Party (PT), the party with the largest number of identifiers in Brazil (Petistas), are both less likely to be classified as generally untrusting and more likely to be generally trusting. In the UK we see a tendency for people who identify with the right-leaning Conservative party or the Liberal Democrats to be more trusting but supporters of the left-leaning Labour party are not significantly more likely to be untrusting. Instead, the untrusting are disproportionately found among those who do not identify with any of the three major political parties in the UK.
Although partisanship may explain only some trust differences, when we asked respondents how favourably they rated political leaders in their country, the results underscore how intertwined political attitudes more broadly are with trust in news. For India, the UK, and Brazil, attitudes towards Narendra Modi, Boris Johnson, and Jair Bolsonaro are strongly correlated with levels of trust in news. In India and the UK, those who evaluated Modi or Johnson favourably were much more likely to be generally trusting towards news than untrusting. This dynamic was reversed in Brazil, where positive evaluations of Bolsonaro were associated with the untrusting instead. The dynamic in Brazil is most similar to the US, where the generally untrusting towards news were also more likely to say they held a favourable opinion of Donald Trump.15 However, unlike in any other country, it is the selectively trusting rather than the generally trusting who held the most unfavourable attitudes towards Trump. These results are largely explained by unusually high favourability ratings towards Trump among the extreme segment of the generally trusting who said they trusted all 15 brands asked about. Among those respondents, 70% had a favourable opinion of the former president.
Differences in trust in news also reflect geographic divides
In addition to demographic characteristics and political preferences, we also found some apparent differences in trust in news depending on where people live. These geographic divides were only partly about regional divides, of which we found some in Brazil, India, and the UK, but not the US, but we also found sometimes wide gaps in trust when comparing people in more rural areas or smaller communities to those in urban centres.
In Brazil, we found that the generally untrusting were somewhat more concentrated in the South, a smaller region characterised by its higher standard of living and higher levels of support for Bolsonaro. Nearly a third of respondents in the southern region (32%) were classified as generally untrusting whereas just 16% were generally trusting. Likewise, people living in the larger Northeast region of Brazil, where there is greater inequality and poverty, tended to be more trusting of news (31% of people in the Northeast were generally trusting versus 20% untrusting). This dynamic was reversed in India, where the Northern region tended to be the place with the highest rates of generally trusting respondents (32%) and the lowest levels of the generally untrusting (22%). Similarly, in the UK, the generally untrusting were less likely to be found in the Greater London area, where 34% were generally trusting of news compared to just 19% generally untrusting. This trend may relate to perceptions of the news media as being London-centric (Newman et al. 2021).
In three of the countries analysed – India, the UK, and the US – there is also a clear relationship between urbanity and trust: in the most rural or smallest places, considerably higher percentages of people are generally untrusting rather than trusting (e.g. 42% vs 23% in India, 40% vs 15% in the US). Likewise, a larger percentage of people living in the most urban places (large cities) have high trust than low trust (e.g. 41% versus 21% in the UK). Trust was not correlated with urbanity in Brazil, although this variable was measured using a different method there.16
2.2 How do the generally untrusting view the world differently?
Next, we examine differences in how the generally untrusting and the generally trusting view the world. We find that in three out of four countries, those who are generally untrusting are also less satisfied with how they think democracy is working in their countries. We also find that social or interpersonal trust may be a factor in contributing to differences in trust in news, with the generally untrusting also much more likely to express scepticism about whether people in general can be trusted. Finally, we also show how the generally untrusting are somewhat less supportive of free expression and somewhat more in favour of government censorship compared to others in their countries. In other words, gaps in trust in news are only partly about the news; they also reflect a broader set of attitudes about how societies are organised and governed.
Those who are less trusting towards news also tend to be less satisfied with democracy
First, we examine the relationship between trust levels and what people think about ‘the way democracy is working’ in their countries. We focus specifically on those who say they are not satisfied and find wide gaps between those who are generally untrusting towards news and those who are generally trusting.
We see the largest gaps in attitudes in the US, perhaps reflecting a greater polarisation in perceptions about the state of democracy there. Almost three in four Americans who are generally untrusting towards news (73%) also say they are not satisfied with how democracy is working. This compares to just 22% of Americans who are generally trusting towards news. We find smaller but similar gaps in the UK and India, where the overall baseline level of satisfaction with democracy is higher.
Brazil is an outlier among the four countries, where trust in news gaps are virtually non-existent on this question. Why does this trend not appear in Brazil? It may reflect already high levels of dissatisfaction across the country. Nearly 60% of Brazilians overall say they are not satisfied with the way democracy is working in their country, and both those who generally trust most news brands and those who generally do not trust most news brands are in broad agreement on that point.
Interpersonal and social trust as drivers of trust in news
We also included several items in the survey intended to evaluate to what extent specific characteristics of individuals themselves – their disposition and outlook on social relations more generally – might contribute to differences in rates of trust and distrust in news. Specifically, we included measures for capturing levels of interpersonal trust, or the degree to which respondents think most other people can be trusted, and social trust, or the degree to which respondents trust specific groups in society. In both cases we find that trust in news is strongly correlated with these other forms of trusting bonds, which suggests that news media itself and the specific journalistic practices of the news in a given media environment may only partly explain differences in trust.
When we look at how trust in news levels differ along lines of interpersonal trust, we find a similar story in India, the UK, and the US.17 To measure interpersonal trust, we asked respondents to weigh which is closer to their view: whether ‘most people can be trusted’ or if ‘you need to be very careful in dealing with people’. This question has long been used by social scientists to measure how trusting people are in general towards people they are unfamiliar with beyond their own close circles of family and friends.18
We find that large majorities of those who are generally untrusting towards news say they need to be very careful in dealing with other people. This contrasts with lower percentages among those who are generally trusting towards most news brands. Differences are most pronounced in the UK and the US, where there is a 27 percentage point gap in both countries versus a 21 percentage point difference in India.
Likewise, when we consider social trust towards specific groups in society, similar patterns appear. In order to assess the salience of different social cleavages, respondents in each country were asked to what extent they trusted people of another race or ethnicity, people of another religion, people from other parts of their country, or people from each of the major political parties. These items varied somewhat from country to country to better fit the relevant cultural contexts, but across the board with only one exception, we find a consistent and similar relationship between trust in news and levels of social trust. People who said they were somewhat or completely trusting of any of these outgroups also tended to be more generally trusting of news brands in their country and vice versa.19
These differences were also pronounced when asking people about how much they trust people who support other political parties. In the US, 30% of those who say they trust Democrats were also generally trusting towards news compared to just 8% among those who say they do not trust Democrats. In general, differences between groups were smallest in Brazil, although the pattern was similar and statistically significant. These results again suggest that one of the drivers of trust in news may be rooted in how trusting individuals are towards other people more generally.
Trust in news differences also extend to attitudes towards censorship
In our survey, we also asked respondents about their opinions on specific democratic norms, especially pertaining to issues of censorship and freedom of speech. More specifically, we asked respondents whether they believe a) people should have the right to criticise their government, b) the media should be able to report the news without government censorship, and c) people should be able to use the internet without government censorship. We found that, in general, majorities in all four countries support these democratic norms. However, we also found that in India, the UK, and the US, those who are more generally untrusting of news tend to be somewhat less in agreement about the importance of these norms relative to the average. In Brazil, the responses show a similar trend but are not statistically significant. Brazil also had the highest baselines for all questions pertaining to censorship.
In the accompanying figure we plot overall agreement with each statement for each country and show how support tends to be lower across the board among those who are generally untrusting towards news. For the question on whether people should have the right to criticise their government, the difference between those in the generally untrusting group and the overall baseline in each country ranges from six to eight percentage points in India, the UK, and the US.
The differences are slightly larger when asking people whether the media should be able to report the news without government censorship. In this case, the difference between those in the generally untrusting groups and the overall percentage who agree is 11 percentage points in both India and the UK, nine percentage points in the US, and four percentage points in Brazil. The differences are again slightly smaller when asking participants whether they believe people should be able to use the internet without government censorship. Here, the difference between people in the generally untrusting group and the average ranges from five to nine percentage points in India, the UK, and the US. Together, these responses suggest that people who are more untrusting of news also tend to value certain democratic norms somewhat less.
2.3 How a lack of interest in news and politics relates to gaps in trust
In closing this section, we consider how interest in news and politics, beyond the other cleavages we have already identified, may also be important factors behind trust in news. We find similar patterns across the four countries, where divides along interest in news and engagement in political life are generally also reflected in gaps between those who are generally untrusting and others.
The generally untrusting use news less frequently and are less interested in politics
We find that in all four countries, less frequent news use consistently correlates with less trust: people who say they access news less often than once a day tend to be more untrusting. The differences are particularly large in the US, where less frequent news users are overwhelmingly concentrated among the generally untrusting (51%), compared to just 23% of those who are generally trusting. Likewise in the UK, the proportion of the generally untrusting who use news less often than once a day is twice as large when compared to the generally trusting (34% compared to 17%). Differences are smaller in India (10% versus 2%) and Brazil (21% versus 18%).
We find even more consistent patterns when examining trust in news in relation to self-reported interest in news and interest in politics. In all four countries, people who are uninterested in news are much more likely to be untrusting of news compared to the generally trusting group. For example, in the UK, those who are uninterested in news make up only 7% of the generally trusting group compared to 27% of the untrusting group. On the flipside, those who are very or extremely interested in news are much more inclined to be in the group that is generally trusting of news. This holds true across all four countries. We find a similar trend when looking at interest in politics: those who are uninterested also tend to be more untrusting of news. We find the largest skew in the US, where the politically uninterested make up 42% of the generally untrusting group and only 17% of the trusting group. The pattern is weakest, but still significant, in Brazil, where people uninterested in politics make up 35% of the untrusting group compared to 32% of the country overall.
The generally untrusting feel less understood by news organisations
People who are generally untrusting of news brands not only tend to be less interested in news and politics, but they also tend to feel less understood by news organisations in general. This holds true across all four countries, although a considerably larger percentage of people in India say they feel news organisations understand them. In all countries, we find much smaller percentages of people in the untrusting group who say they feel that news organisations ‘understand people like me’ compared to the overall baselines in each country. In the US, the percentage is as low as 16%, compared to 49% overall and 81% of those in the generally trusting group. Among people who trust no brands at all or trust all brands, the gap is even wider (10% versus 88%).
We also see a consistent pattern in the relationship between feeling understood by news organisations and political interest. In all four countries, the less interested people say they are in politics, the less they feel understood by news organisations. These differences are greatest in the UK, where 73% of people who are extremely interested in politics feel understood by news organisations, compared to only 38% of people who are not interested in politics at all. Likewise in the US, 65% of those who are extremely interested in politics feel understood by news organisations, compared to only 24% of people who are not interested in politics at all. The differences between the extremes in political interest are smallest in Brazil, where the percentage of people who feel interested in politics overall is lowest of all four countries (49% of those extremely interested in politics feel understood versus 38% of those not at all interested). These findings suggest that the orientation of news organisations towards political subjects contributes to people with low levels of political interest feeling particularly misunderstood.20
Ideas about whether or not news brands understand people like them also relate to whether people think most news organisations are genuinely interested in engaging with the public. While majorities in India, the UK, and the US and 33% in Brazil believe news organisations do in fact want to hear from the public, the share that agrees among the generally untrusting is far lower. In the US, this gap is once again particularly stark, with just 15% of the generally untrusting agreeing that news organisations genuinely want to hear from the public. This compares to 50% of Americans overall and 80% of those who are generally trusting of news. In Brazil, where only 33% of people overall think that news organisations are interested in hearing from the public, just 19% of those who are generally untrusting agree, compared to 42% of those who are generally trusting. Even in India, where the overall percentage of people who think news organisations are interested in hearing from the public is highest, the differences between the generally untrusting (47%), the country overall (65%), and the generally trusting (84%) are considerable.
In the next section we elaborate further on how to interpret these results; while many of the generally untrusting have specific complaints about how journalism is practised in their countries, they are not necessarily highly engaged media critics. Instead, unlike the selectively trusting, many of the generally untrusting are relatively indifferent about the specifics of how news is practised. As we show elsewhere in this report, one explanation for this divergence in views is that the generally untrusting have also had fewer experiences interacting with journalists or engaging with brands and are simply less familiar with what differentiates one source from another, especially online. Winning over these mostly indifferent audiences may require different strategies than building back trust with other, more opinionated segments of the public.
3. Indifference as an underappreciated driver of low trust ↑
Our primary finding in this section is that people who are generally untrusting towards news pay considerably less attention to newsgathering and editorial practices relative to people who are more trusting. We find a consistent gap along lines of trust across all four countries in how important different journalistic practices tend to be when people weigh whether to get news from one source versus another. Although people may not always recognise what is and is not important to them when making such decisions – after all, many people are exposed to news incidentally online without necessarily seeking it out (Fletcher and Nielsen 2018; Thorson 2020) – how people think about and explain their own media selection behaviours is revealing in its own right. We find that the generally untrusting, in contrast to other groups, tend to be less opinionated about how reporters, editors, and publishers should do their jobs and are less likely than others to say they care about even some broadly popular ideas about how news should work, such as the watchdog role of the press or how entertaining or engaging news should be. We think this points to indifference towards news as a major factor underlying the lack of trust seen among the generally untrusting.
3.1 The generally untrusting say they are less concerned about editorial practices or disclosure of journalists’ backgrounds
In a previous report (Toff et al. 2021), we observed that many people do not hold clear notions about what individual news brands stand for or how to evaluate the reporting they produce. Instead, when navigating the media landscape, people often draw on shortcuts informed by past experiences, word-of-mouth, or general impressions held about news media more generally. These ideas about news, and people’s self-reflections about how they use news, may not always be accurate and are often influenced by messages from elites, but regardless they can play a significant role in shaping future behaviour, as previous journalism research has argued (e.g. Nelson and Lewis 2021).
We included several items in our survey designed to investigate the kinds of things people thought were important when they made choices between sources they encountered. We asked about editorial practices, including the importance of knowing about how journalists find and choose sources for their stories as well as why stories are chosen in the first place. We also asked about the importance of financial transparency, a subject which has previously garnered a great deal of attention from both practitioners and scholars (Porlezza and Splendore 2016; Rosenstiel et al. 2016). In addition, we included a series of similar questions gauging how much respondents cared to know about the backgrounds of individual journalists: their subject matter expertise and their political preferences, as well as their gender, race, or religious background. Respondents were asked to state how important each of these factors was to them personally on a four-point scale from ‘not important at all’ to ‘extremely important’.
We find an almost uniform pattern in all four countries: the generally untrusting are less concerned about either editorial practices or journalists’ backgrounds. We do not think this is because transparency around such matters does not matter. Instead, the generally untrusting tend to be much less engaged with journalism altogether. Many may default to a lack of trust out of indifference and uncertainty.
The generally untrusting say they pay relatively less attention to editorial practices
The pattern we describe is replicated across all the questions we asked concerning the importance of different factors involved in how news is made. The gap between the generally untrusting and generally trusting is often 20 percentage points or more.
Take, for example, differences in the percentage in each country who say they think it is important to know how news outlets find and choose sources for their stories. There are both country-level differences – respondents in Brazil and India compared to the US and the UK are typically more likely to say that knowing such matters is important to them – but there are also consistent patterns within countries. About two-thirds of generally untrusting respondents in Brazil (68%) and India (67%) say knowing about how news organisations make sourcing decisions is very or extremely important to them. Less than half of the generally untrusting in the US (46%) and the UK (41%) say the same.
A similar pattern is apparent when we look at other indicators of what people pay attention to when they evaluate news. When asked whether it was important for respondents to know why stories were chosen in the first place, the gap between generally untrusting and generally trusting respondents in whether knowing this information was very or extremely important to them was 24 percentage points on average across the four countries.
We find much the same dynamic when it comes to disclosure about news organisations’ funding. Although the gap between the trusting and untrusting groups is slightly smaller in Brazil, the UK, and the US, in the case of India the generally trusting are much more likely (by 29 percentage points) to say that knowing about news organisations’ funding is very or extremely important relative to the untrusting group.
One factor that partially contributes to these disparities is that the generally untrusting are also somewhat more likely than other groups to say they haven’t thought about this subject before. Overall about 15% in the US and the UK and 6% in India said they had not previously thought about this question about news organisations’ funding.21 Among those who did, they were significantly more likely to be among the generally untrusting. In the US, for example, those who said they hadn’t thought about how much news organisations’ funding mattered to them were nearly twice as likely to be in the generally untrusting group versus the generally trusting group.
Knowing more about who reports the news tends to be viewed as relatively less important – except when it comes to politics
Next, we examined whether journalists’ backgrounds matter to people when deciding between news sources. We first asked respondents whether it is important to know about what expertise reporters have on the subjects they are covering. This was the one question we asked where a majority in all four countries said it was very or extremely important to know. However, we again find that indifference to this subject is greatest among the generally untrusting, often by a 20–30 percentage point margin. In the US and the UK, fewer than half of the untrusting say they think it is very important to know these details.
When we asked respondents about several other characteristics involving journalists’ backgrounds, we find smaller percentages in all countries report a strong desire to know reporters’ gender, religion, or race (or caste in India). The one background characteristic there is wider agreement about is the partisan identity or political leanings of journalists. Overall, more than twice as many people say it is very or extremely important to know the political leanings of journalists when choosing between news sources, compared to gender, religion, or race.
It is not, however, the generally untrusting who are driving the higher rates of interest in knowing the political affiliations of journalists. We see a similar trend here in the gap between the generally trusting and untrusting groups; however, the differences on this question are much smaller compared to others – in some cases only a few percentage points. Furthermore, it is instead the generally trusting, who tend to be more engaged and interested in politics, that are most interested in knowing about the political leanings of journalists who report the news. It is worth noting, but perhaps not surprising, that interest in knowing about the political leanings of journalists is strongly correlated with interest in politics. More than two-thirds of people in the US and the UK who say they are ‘extremely’ interested in politics say they are very or extremely interested in knowing the political leanings of the people who report the news. In Brazil and India, roughly 80% say the same.
When looked at by country, both India and Brazil stand out given that more than six in ten people overall say they think it is very or extremely important to know the political leanings of journalists. This is 20 percentage points higher than the rate in the US, where partisan polarisation has been the focus of a great deal of attention from practitioners and scholars (e.g. Jurkowitz et al. 2020) and where a tradition of nonpartisan neutrality in news may be a relatively more ingrained part of the public’s expectations about journalists than elsewhere.
Somewhat higher levels of interest in knowing about the political leanings of journalists does not necessarily mean people reject the notion that news should be impartial. In fact, when we asked respondents to what extent they agreed that journalists should ‘separate facts from opinions when covering the news’, more than 80% agreed in all four countries and as many as seven in ten of the generally untrusting everywhere as well. However, when we asked respondents whether they thought journalists should take the proactive step of disclosing their political leanings, there was much less agreement. While more people agreed than disagreed, only in India was the share of respondents who agreed an overwhelming majority (69%).
3.2 The public demands a lot from the news, but the least trusting are also the least demanding
In a separate part of the survey, respondents were also asked several other questions designed to assess what other characteristics were important to them when they make evaluations about news. These include expectations about whether news organisations should serve as independent watchdogs policing the conduct of those who wield power, but also other functions of the news, such as providing analysis of complex matters or communicating in ways that are easy to understand. These kinds of questions involved ideas about what journalism should be and what reporters should do.
On these items, we also find across the board that the more generally untrusting are also less likely than others to say they care about even some broadly popular ideas about how news should work.
Most agree that journalists should hold power to account but the generally untrusting are more indifferent
Of the statements we measured agreement on that involved expectations about the roles journalism should play in society, two focus on politics: we asked respondents whether they agree or disagree that journalists should investigate claims and statements by the government and whether journalists should hold politicians accountable.
We find widespread agreement on both of these measures. More than three-quarters in all countries agreed with the statements, but we also find significant gaps by levels of trust. We find that those who are untrusting towards news are less likely to agree with both statements. The generally trusting, in contrast, overwhelmingly embrace the notion of the press as a watchdog. Brazilian respondents in particular show the highest levels of support for these statements, and the differences between the generally untrusting and trusting there are also considerably smaller.
The widest trust gaps emerge on questions about the content and presentation of news
We next look at what respondents think about the way news is delivered and the kinds of topics news outlets tend to focus on. We examine how respondents vary in whether they think news organisations should make news entertaining, whether they should provide analysis and interpretation of complex problems, whether they should communicate in ways that are easy to understand, and whether news should highlight good things that happen in the world, not just the bad. We also ask respondents whether they think news organisations should give ordinary people a chance to express their views on public affairs.
We find some of the widest gaps in trust in news on whether news outlets should make news more entertaining. Our previous qualitative data (Toff et al. 2021) revealed that some respondents, especially in India, said they consumed TV news especially for its entertainment value. We were interested in understanding how widespread such views might be. Agreement is highest in India, where about two-thirds of respondents say journalists should make news entertaining so people pay attention to it. Agreement is lowest in the US (41%), but we also find a striking divide between the generally untrusting and trusting, with far fewer of those who are untrusting saying they agree news should be made entertaining. In the US, this gap is 43 percentage points but just 16 percentage points in Brazil. These differences are intriguing since the generally untrusting also say they consume news less frequently. It is as though, in rejecting such notions, the generally untrusting are saying making news more entertaining is unlikely to get them to change their own behaviours.22
When it comes to other aspects of news content, we find that the generally untrusting are also generally more indifferent about editorial decisions. While majorities agree with all four statements that news should provide a platform for ordinary people to express their views, provide analysis of complex problems, communicate in ways that are easy to understand, and focus on highlighting good things in the world, the generally untrusting are consistently less likely to agree across all countries.
What kinds of heuristics do people use when they navigate the media environment?
Lastly, we examined three questions designed to assess what kinds of heuristics, or cues, people said they relied on when making choices about news they encounter in daily life. We asked people to what extent they agreed with three statements: (a) the quality of a news outlet’s grammar or language is a good indicator of whether it is trustworthy; (b) I rely on friends and family when determining what news I can trust; and (c) I am more likely to trust news that includes numbers, graphs, or statistics as evidence.
We find that, overall, a higher share of respondents in India and Brazil (68% and 61%) say they rely on grammar or language; the share is lower in the UK and the US (56% and 48%). However, in all countries there is variation by the level of trust, with those who are generally untrusting saying they rely on these indicators to a far lesser extent. Respondents in India in the generally trusting group (as opposed to the untrusting) were the most likely to say they rely on this heuristic (86%).
Among other differences, respondents in India were also considerably more likely to say they relied on friends or family for help with determining what news they can trust (61%), although in other countries only about a quarter of respondents said the same.
About half of those in the US and the UK said they interpreted the use of numbers, graphs, or statistics as a sign that news was more trustworthy, whereas rates agreeing with this statement were higher in Brazil (63%) and India (78%). But again, like the other items we asked about, the generally untrusting were considerably less likely to agree.
In sum, these patterns suggest that on a range of different measures, gaps in trust are not necessarily driven by unmet expectations about the performance of news or the generally untrusting paying particular attention to certain facets of coverage compared to other groups. Instead, we find a consistent pattern where the least trusting in each country are simply less interested in news and therefore tend to say they care less about different factors governing the way it is produced and disseminated.
4. Many are uncertain about the practices of journalism ↑
Past research has indicated that knowledge about journalistic practices is often low (Media Insight Project 2018). Experiences directly interacting with journalists are also often rare (van der Wurff and Schönbach 2014). As we show in this section, this lack of contact and familiarity with even basic journalistic concepts is often accompanied by low expectations about how news organisations conduct themselves. To varying degrees, this dynamic is present in all four countries; people who are less knowledgeable about how journalism works also tend to have less interest in learning more about it.
4.1 How interacting with journalists is associated with more trust
Roughly three-quarters of respondents in the UK and the US and just under half the sample in India said they had never known a journalist personally, been interviewed for a news story before, or contacted a journalist with ideas or feedback. (These questions were not asked in Brazil.) Respondents from India were somewhat more likely to report all three kinds of interactions compared to the other countries – differences which may be attributable to country-specific survey response styles or interpretations of questions (Harzing 2006) – but even so, two-thirds of respondents in India also said they had never known a journalist and had fewer other kinds of interactions.
When we consider the relationship between interacting with journalists and trust in news, we find that people who report having experienced any of these interactions tend to be overrepresented among the generally trusting segments of the public in all three countries. In India, 71% of those who are generally trusting say they have experienced at least one of the three forms of contact with journalists we asked about. This compares to 47% of the generally untrusting. In the UK, just 17% of those with generally untrusting attitudes towards news brands say they have interacted with a journalist, compared to 31% of generally trusting respondents. The pattern is similar in the US, where 20% of those who are generally untrusting say they have experienced at least one of the interactions we asked about but 42% of those who are generally trusting say the same.
Familiarity with basic terminology of news varies greatly but is highest among the most trusting
While it is difficult to assess actual knowledge about newsgathering practices, we included a series of questions that asked respondents to report how familiar they were with some common terms and concepts used to describe types of news content. Many of these items were adapted from a previous survey conducted in the US by the Media Insight Project (2018).
We find large majorities in three countries said they were familiar with what the term ‘breaking news’ means; eight in ten in India and seven in ten in the UK and the US said they were ‘completely’ or ‘very’ familiar with the term. However, in Brazil the term was much less widely recognised – just 45% reported such a high level of familiarity. In comparison, a slightly higher percentage in Brazil (48%) said they were completely or very familiar with the difference between a reporter and a commentator.
Familiarity levels with other terms also varied by country – a reflection of important differences in journalistic cultures. In Brazil, the UK, and the US, the least familiar concept was the difference between an editorial and a news story, where less than half said they were very or completely familiar with it (38%, 36%, and 49%, respectively). The percentage was again highest in India (63%). In general, Brazilian respondents expressed lower levels of familiarity with any of the terms and concepts, which may be attributable to a more limited tradition of criticism and media literacy in the country, at least in comparison with the UK or the US (Marques and Miola 2021).23
Next, in line with results discussed in previous sections, we find that those who are generally trusting of news also say they are more familiar with these journalistic terms and concepts compared to those in the generally untrusting segments of the public. This pattern is again consistent in all four countries. In Brazil, for example, just 36% of those who are generally untrusting say they are familiar with the term ‘breaking news’, compared to half of those who are generally trusting. Similar or larger gaps exist in the other countries. In another example, 61% of Americans who are generally trusting of most news brands say they are very or completely familiar with the difference between an editorial and a news story, but only 36% of those who are generally untrusting in the US say the same.
It is impossible to assess whether gaps in familiarity with journalistic terms and concepts cause the apparent differences in trust observed. Of course, differences may also reflect other aspects that are aligned with trust, especially education and other inequalities, as described in section 2 of this report. Indeed, respondents who interacted with journalists in any of the three ways we asked about also tended to be more familiar with journalistic terms and concepts in general.24 Nonetheless, the stability of this pattern across countries does suggest at least an association between knowledge of news practices and trust.25
4.2 Most default to negative ideas about how journalism is practised
Our results so far indicate that many respondents have limited familiarity with journalistic concepts and that this lack of familiarity tends to be associated with a lack of trust. Another relevant trait present in all countries, even if in different degrees, is the scepticism many people display towards news organisations.
When asked how they think journalists do their jobs, a significant proportion express agreement with fairly negative ideas. While most respondents think journalists sometimes double-check facts with multiple sources (70% in Brazil, 72% in India, 66% in the UK, and 63% in the US), that still leaves nearly one-third in each country who say they think journalists do so only very rarely or never.
While a significant minority in the four countries hold negative views about whether journalists double-check their facts, the generally untrusting in each country are even less likely to say they think this happens ‘very often’. In every country, the more trusting a person is, the more positive their attitudes about journalists’ practices. For example, 46% of respondents from India who are generally trusting in news think journalists double-check their information very often, while 21% of those generally untrusting say the same.
Respondents also express suspicion when they evaluate whether news organisations are willing or not to admit their mistakes. In all countries, majorities of respondents say they think news organisations try to cover up their mistakes, with the highest proportion in Brazil, where more than three-quarters said so. Respondents from India and the US are more divided, with the percentage saying they thought news organisations were willing to admit their mistakes topping 40%.
Those who are more generally trusting towards news also tend to say news organisations are willing to admit their mistakes, with significant gaps in the US, the UK, and India by levels of trust. Nearly one-third of Indian respondents (32%) who are generally untrusting, 19% of those in the UK, and 14% of those in the US say news organisations are willing to admit their mistakes. In these cases, scepticism is not widespread across all respondents, but the gap between people with different levels of trust also indicates how people in the same country perceive news organisations in very distinct ways. The pattern is the same for Brazil, but very few hold positive views about whether news organisations are willing to admit their mistakes.
We also asked a follow-up question in India, the UK, and the US about perceptions not only about news in general in their country, but also specifically of the news they themselves use. As expected, people have more positive evaluations when thinking about news they use. In India, for example, 60% of respondents say news organisations they use themselves are willing to admit their mistakes compared to 45% who say the same for news in general. We find similar differences in the UK and the US. This may be another indication of how familiarity is a central element for people when making decisions about news sources they trust.
Many think news organisations seek to manipulate the public with their own agendas
Negative default views about the press become even more evident when we asked about other journalistic practices. We find that as many as four in ten respondents say they think journalists ‘very often’ try to manipulate the public. The level is highest in Brazil (43%) and lowest in the UK (30%). We also find relatively large minorities of respondents who similarly believe that news organisations ‘very often’ intentionally try to provoke in order to get attention for themselves, allow their personal opinions to influence their reporting, and write articles before they learn the facts about events. More than one-third of respondents in Brazil (36%) and India (37%) believe journalists very often get paid by their sources. However, the percentage who believe the same in the UK (25%) and the US (27%) is not much lower.
When we compare these results by levels of trust in news, we find similar gaps as elsewhere between the generally trusting and untrusting; however, cynicism about journalistic practices occurs even among those who are otherwise generally trusting. That is, people at all trust levels often hold negative perceptions about how journalists do their work. Nonetheless, more positive perceptions are more common among those who are generally untrusting. By a nearly two-to-one margin, Brazilians and Americans who are generally untrusting are more likely to believe journalists try to manipulate the public compared to the generally trusting (64% versus 33% in Brazil and 59% versus 25% in the US). Smaller gaps are also present in India (46% versus 40%) and the UK (41% versus 24%).26
4.3 The link between brand use and trust
Why might people hold such negative views about how journalists operate and yet hold somewhat trusting views about the information reported in the news media (as we show in section 1 of this report)? One explanation may be that people hold low expectations about how journalism works when they think about reporters and broadcasters in the abstract. Some may have a more positive outlook when thinking about specific brands they are most familiar with.
In fact, we do find evidence across all four countries that trust tends to be highest among respondents who say they used a particular news brand during the previous week. In some cases, those who used brands on social media hold even higher levels of trust compared to those who accessed them via other modes, but these differences pale in comparison to the gap between non-use and use. Those who say they did not use a given brand during the previous week tend to rate that brand significantly lower.
This pattern does vary to some degree depending on the brand, but the trend is surprisingly consistent. In Brazil, for example, the average level of trust among those who say they used Globo is almost twice the level of those who say they did not use it during the previous week. This gap amounts to 1.3 points on a four-point scale, the largest for any brand across all four countries. This disparity may be a consequence of President Bolsonaro’s antagonising rhetoric,27 as well as the brand’s history of conflicts with left-wing partisans.28 Significant gaps in trust related to use are present for other brands as well, such as Terça Livre in Brazil (1.1), Republic TV in India (0.6), Fox News in the US (1.0), and the Sun (0.8) and the Daily Mirror (0.7) in the UK. BBC News (0.6) is among the brands with the lowest gaps in trust between those who use it and those who do not consume it frequently.
Many say they revert to being automatically sceptical in the absence of direct experiences using many news brands
In the third section of this report, we discussed some of the heuristics people said they used when they navigated the news environment and made decisions about which sources to trust. Here we focus on several additional strategies people said they employed. Three stand out in particular: (a) large majorities in all countries said they thought it was important to be sceptical towards all sources of news unless you have yourself verified what has been reported with multiple sources, (b) almost as many said they tended to trust news organisations more that they were more familiar with, and (c) more than half, and as many as three-quarters in India, said they were less trusting of all news sources when they encountered ‘different accounts of what’s going on’. Overall levels of agreement with these statements varied by country, but these ideas were widely embraced.
These sentiments are very much in line with our previous findings (Toff et al. 2021), which were based on small samples interviewed qualitatively. While highly engaged segments of the public in each of these countries hold specific, detailed views (positive and negative) towards a large number of news brands, many others rely on more impressionistic methods to differentiate between sources. When they encounter disagreement between them, as is common in the contemporary information ecosystem, most people see it as a reason to be sceptical towards all news, especially when unfamiliar with most news organisations’ reporting practices or other editorial standards. Few have the time or inclination to investigate such details on their own.
The one statement we asked about that was not embraced by a majority of respondents, except in India, was the idea that trustworthiness could be detected on the basis of how a news source looks – that is, how it presents itself visually. Only 28% of respondents in Brazil and 32% of people in the US and the UK said they agreed, although the level was roughly twice as high in India (61%), where the production values of news organisations may vary more widely than elsewhere. It is possible – likely even – that responses to this question are subject to social desirability biases; that is, many more people may in fact use these cues than are willing to admit to it, because they believe it is an unsophisticated way of differentiating between sources.
Interestingly, those who are generally trusting towards news brands in all four countries also say they are more cautious when evaluating news sources. For example, as many as 70% of those who are generally trusting in the UK say it is important to be sceptical of all news sources, compared to 59% of those who are generally untrusting. The generally trusting also tend to say they pay more attention, rather than less, to how familiar a brand is and even its visual presentation and design compared to other groups. For example, 89% of Indian respondents who are generally trusting agree that the more familiar they are with a news organisation, the more likely they are to trust its reporting, whereas just 60% of the generally untrusting say the same.
These results again corroborate findings from our previous report (Toff et al. 2021), which suggested that familiarity with brands often played a key role in shaping what people think about a news organisation’s news content. In this case, we show that familiarity not only makes it easier for people to pick which sources to use, it can also be a driver of trust.
In short, our results offer additional evidence that the generally untrusting as a group are somewhat less confident in their ability to differentiate between sources, and that may contribute to a lack of trust overall. The untrusting not only say they pay less attention to journalists’ backgrounds and the editorial practices individual news outlets embrace, they also put less stock in the way brands present themselves or whether familiarity is even a useful indicator of trustworthiness. On average they trust few or no brands, not because they are particularly discerning but because they are less knowledgeable about what separates one brand from the next and may lack the motivation or interest to find out.
This report details findings from surveys conducted in May and June 2021 in Brazil, India, the UK, and the US. We classify the public in each of these countries as ‘generally untrusting’, ‘selectively trusting’, or ‘generally trusting’ towards specific news brands in their countries and show how differences in trust in news relate not only to specific attitudes about journalism and how it is practised, but also to other institutions, especially political ones. We further show that the generally untrusting towards news tend to be less educated, older, and disconnected from urban centres – patterns that are consistent in most countries. Unlike more vocal and visible critics voicing specific real or perceived grievances, they are often more indifferent towards journalism generally. They say they pay less attention to the backgrounds and expertise of reporters and are somewhat less concerned with how news organisations should conduct themselves. Most are less familiar with basic journalistic terms and concepts, do not differentiate much between brands, and tend to lack interest in the news they encounter.
The challenge news media and journalists face if they want to earn the trust of this part of the public is not necessarily overcoming entrenched hostility but addressing something perhaps even more insidious: indifference. The issue here is not so much a deep-seated difference of opinion over the value of news so much as a relative lack of interest in what news is or ought to be. This indifference may be rooted at least in part in more limited experience using news or interacting with journalists, but also in low expectations about how news can and should be practised.
Implications of our findings
While perhaps distressing to professional journalists and other practitioners in the news media – who might prefer that all news audiences appreciate their organisations’ contributions to the greater good and acknowledge what is unique about each brand’s offerings compared to competitors – our findings also offer insights into what can be done to build trust with different groups.
Leaving aside the generally trusting who appreciate most brands, and the selectively trusting who have at least some brands they value (even as they are sometimes vocal critics of others), we focus here specifically on the generally untrusting, who constitute about a quarter of respondents in each country and often have more limited and peripheral contact with news.
First, the general challenge for journalists and the news media here is how to communicate to this group (a) the ways in which journalism is different from all the other types of information and forms of self-interested and strategic communication so abundantly available, and (b) why it is valuable to them, individually, in addition to the value it holds for others and society more broadly. Journalists and news media may think that the unique value proposition of independent, professionally produced news is obvious. To much of the public, it isn’t. Professionals might think that the work speaks for itself. It clearly doesn’t. This is a communications challenge.
Second, the specific challenge for individual brands that want to engage with and build trust among the generally untrusting is that they need to (a) reach a group unlikely to come directly to news sites or news apps, which may mean engaging with platforms that publishers are wary of, and (b) demonstrate enough value to build a minimum of habit and loyalty so that people can experience and see for themselves the value of the news and journalism on offer. It is important to be clear that many for-profit news media may, frankly, not want to do this because the commercial returns of focusing on competing for the attention of affluent and already engaged people are likely to be greater. But popular media may see this as a different niche setting them apart from upmarket competitors, and public service media in some countries are required to provide a universal service, so they will have to think about this group. This is a marketing challenge.
Third, the editorial challenge for newsrooms involves a potential reordering of some journalistic priorities to produce work that is perceived as valuable and deserving of engagement from parts of the public who otherwise have limited interest in most existing news. That is, how can news outlets persuade the generally untrusting that their organisations are, in fact, willing to admit their mistakes, genuinely want to hear from the public, and fundamentally do understand people like them? How can news outlets better demonstrate a desire to report on issues that extend beyond the intricacies of daily politics, or cover the good things happening in the world, not only all that is negative and depressing? The generally untrusting judge news media harshly on all of these characteristics, but so too do many others. This is a creative challenge that goes beyond communicating and marketing and involves making the editorial product itself better – or at least more inclusive.
These are initiatives that the news media, or individual news media, could take, and that our research suggests could help engage and build trust with a large part of the public that is currently somewhat indifferent towards the news. This does not mean that such steps on their own are more important than other more structural factors that are beyond the control of any journalist or editor, whether social and political (changes in institutional or interpersonal trust) or driven by audience preferences and technological developments (the move to a more digital, mobile, and platform-dominated media environment). Ultimately, these trends shape trust more than anything any single journalist or news organisation can do, but that does not mean that there is nothing to be done in response.
Further enhancing trust among the generally trusting or the selectively trusting may perhaps be achieved by other means. After all, different groups will require different approaches. Recent years have seen much professional and scholarly interest in various forms of deep engagement with news as possible solutions to rising levels of distrust among the public (see, for example, Ferrucci et al. 2020; Wenzel 2020). These are promising for some specific audiences, but given that our analysis shows that many of the generally untrusting are relatively disconnected from news and more indifferent towards matters of inclusion or transparency around professional practices, attempts to build trust through initiatives like listening sessions and other formats that require extensive commitments from audiences or outlays of time, while potentially effective for some groups, are unlikely to work for the broad cross-section of people who are the least trusting in general. While there are certainly some communities starved of attention from journalists where such efforts may be welcomed as long overdue, we worry that many of the least trusting are also likely to be the least interested in such approaches, which risk primarily engaging the already engaged.
More broadly, our findings expand on what is known about how attitudes towards news are structured. Deep-seated hostility towards the press (as illustrated on the political right in the US) can reinforce itself through the combination of elite cues from trusted political leaders, negative coverage of news media from other news media, and a daily life spent in communities where many others are sceptical of, or indifferent towards, news media. On the other hand, the reverse can also occur when respected politicians highlight the value of independent professional journalism, news media demonstrate that value, and people are part of communities where many follow and value news. Between this ‘vicious circle’ and ‘virtuous circle’ lies what we might call a ‘vacant circle’, characterised by little interest in politics, limited engagement with news, and little social reinforcement encouraging news use. It is important that news media and those who care about the future of journalism, including public trust in news, seek ways to address these challenges now, as the continued move to a more digital, mobile, and platform-dominated environment may well accentuate many of the challenges identified here. When news captures only a small share of attention and many do not seek it out, and where only limited contextual cues are provided online to set brands apart from other types of information, forging enduring trusting relationships among those least interested in news is as important as it is daunting.
Focus of future research
This is the third published report of our Trust in News Project, and in the next two years we aim to build on our previous research in several ways. We also hope this research inspires other researchers and news organisations to pursue additional inquiries beyond investigating the questions we ourselves are able to address.
First, much of this report focuses especially on the generally untrusting in relation to other groups, and we think investigating variation within the other two groups – especially the selectively trusting – may be worthy of additional attention. Developing and cultivating trust with the selectively trusting – who already trust some brands but not others – is in some ways a less formidable task than connecting with the untrusting, and it may be attractive to individual news organisations who may be more focused on trust in their own brand than trust in news in general.
Second, in our analysis of the generally untrusting we have sometimes oversimplified our case in suggesting that this group is entirely indifferent towards news. In truth, this group also holds often very low expectations about journalistic practices. They are more likely to believe news organisations seek to manipulate the public and try to cover up their mistakes rather than come clean about where they got it wrong. We think these kinds of cynical views also deserve more attention in future research. Where do such notions come from, how do they take root, and what might it take to change people’s minds about whether such practices are the norm or the exception?
Finally, while we believe there is much to learn from surveying and interviewing news audiences, such methods have significant limitations when it comes to understanding the causal factors that underpin attitudes of trust and distrust. We believe what is needed are more experiments, both in the lab and in the field, that test the effectiveness of different interventions around communicating some of the messages we describe above. Building and sustaining trust likely requires different approaches for different audience segments. By testing how different groups respond to different messages, news outlets can develop an evidence-based strategy for how best to move forward.
Appendix: Methodology ↑
This study has been designed and commissioned by the RISJ’s Trust in News Project to better understand the drivers of trust and distrust towards news media in Brazil, India, the UK, and the US. Survey fieldwork was conducted by a combination of three research firms using a mix of survey modes in May–June 2021. In Brazil, surveys were fielded over the telephone by Instituto Datafolha, the polling operation for Grupo Folha. Elsewhere, surveys were fielded online by Internet Research Bureau (IRB) in India and by Kantar in the UK and the US. The Trust in News Project team independently compiled the data from these firms and analysed the results.
- Nationally representative samples were assembled in each country. In Brazil, respondents were randomly selected on the basis of a list of mobile phone number prefixes and weighted to census targets for age, gender, region, and education. Elsewhere, respondents were selected from online panels using quotas for age, gender, region, social grade (the UK), race and ethnicity (the US), and education (the US). The data in the UK and the US were also weighted to targets based on census data.
- In Brazil, Datafolha supplied a sample sourced from their own proprietary database. In India, the IRB recruited 80% of its sample using its own panel and 20% using third-party offline recruiters on WhatsApp or telephone to boost participation in smaller Tier 2 and 3 cities and among non-English speakers. In the UK and the US, Kantar recruited respondents using the Kantar Profiles panel network.
- Surveys were conducted primarily in Portuguese (Brazil) and English (India, the UK, and the US). In the US, the survey was also provided in Spanish. In India, the questionnaire was also translated into Hindi, Tamil, Telugu, Marathi, Bengali, and Kannada.
- The survey questionnaire (13 minutes in length on average online) was generally the same across the four countries; however, some changes were necessary due to translation issues or mode differences that arise when completing the survey over the telephone instead of online. In Brazil, several items were excluded from the survey in an effort to keep the overall length of the questionnaire to less than 20 minutes.
- It is important to note that some of the questions included in the survey rely on recall, which is often imperfect or subject to biases. We have tried to mitigate these risks through careful questionnaire design and testing. On the other hand, surveys can be a good way of capturing fragmented media consumption across platforms (e.g. social media, messaging apps, and websites) and tracking activities and changes over time.
- View the technical appendices for each surveyed country.
- Daniller, A., Allen, D., Tallevi, A., Mutz, D. C. 2017. ‘Measuring Trust in the Press in a Changing Media Environment’, Communication Methods and Measures 11(1), 76–85.
- Delhey, J., Newton, K., Welzel, C. 2011. ‘How General is Trust in “Most People”? Solving the Radius of Trust Problem’, American Sociological Review 76(5), 786–807.
- Ferrucci, P., Nelson, J. L., Davis, M. P. 2020. ‘From “Public Journalism” to “Engaged Journalism”: Imagined Audiences and Denigrating Discourse’, International Journal of Communication 14, 19.
- Fletcher, R. 2020. Trust Will Get Worse Before It Gets Better. Oxford: Reuters Institute for the Study of Journalism.
- Fletcher, R., Nielsen, R. K. 2018. ‘Are People Incidentally Exposed to News on Social Media? A Comparative Analysis’, New Media & Society 20(7), 2450–68.
- Gursky, J., Woolley, S. 2021. Countering Disinformation and Protecting Democratic Communication on Encrypted Messaging Applications. The Brookings Institution.
- Hanitzsch, T., Van Dalen, A., Steindl, N. 2018. ‘Caught in the Nexus: A Comparative and Longitudinal Analysis of Public Trust in the Press’, The International Journal of Press/Politics, 23(1), 3–23.
- Harzing, A. W. 2006. ‘Response Styles in Cross-National Survey Research: A 26-Country Study’, International Journal of Cross Cultural Management 6(2), 243–66.
- Jurkowitz, M., Mitchell, A., Shearer, E., Walker, M. 2020. U.S. Media Polarization and the 2020 Election: A Nation Divided. Pew Research Center, 24 Jan.
- Kalogeropoulos, A., Fletcher, R., Nielsen, R. K. 2019. ‘News Brand Attribution in Distributed Environments: Do People Know Where They Get Their News?’ New Media & Society 21(3), 583–601..
- Ladd, J. M. 2011. Why Americans Hate the Media and How it Matters. Princeton, NJ: Princeton University Press.
- Marques, F. P. J., Miola, E. 2021. ‘Key Concepts, Dilemmas, and Trends in Political Communication: A Literature Review Considering the Brazilian Landscape’, Annals of the International Communication Association. doi: 10.1080/23808985.2021.1945479
- Media Insight Project. 2018. Americans and the News Media: What They Do – and Don’t – Understand About Each Other. Arlington, VA: American Press Institute.
- Nelson, J. L., Lewis, S. C. 2021. ‘Only “Sheep” Trust Journalists? How Citizens’ Self-Perceptions Shape their Approach to News’, New Media & Society. doi: 14614448211018160.
- Newman, N., Fletcher, R., Schulz, A., Andi, S., Robertson, C. T., Nielsen, R. K. 2021. Reuters Institute Digital News Report 2021. Oxford: Reuters Institute for the Study of Journalism.
- Norris, P. 2011. Democratic Deficit: Critical Citizens Revisited. Cambridge: Cambridge University Press.
- Porlezza, C., Splendore, S. 2016. ‘Accountability and Transparency of Entrepreneurial Journalism: Unresolved Ethical Issues in Crowdfunded Journalism Projects’, Journalism Practice 10(2), 196–216.
- Rosenstiel, T., Buzenber, W., Connelly, M., Loker, K. 2016. How News Organizations Handle Disclosure and Transparency of Funding. American Press Institute.
- Schiffrin, A. 2019. Credibility and Trust in Journalism. Oxford Research Encyclopedia of Communication.
- Strauß, N., Huber, B., Gil de Zúñiga, H. 2020. ‘“Yes, I Saw It – But Didn’t Read It …” A Cross-Country Study, Exploring Relationships between Incidental News Exposure and News Use across Platforms’, Digital Journalism, 8(9), 1181–205.
- Strömbäck, J., Tsfati, Y., Boomgaarden, H., Damstra, A., Lindgren, E., Vliegenthart, R., Lindholm, T. 2020. ‘News Media Trust and Its Impact on Media Use: Towards a Framework for Future Research’, Annals of the International Communication Association, 44(2), 139–56.
- Thorson, K. 2020. ‘Attracting the News: Algorithms, Platforms, and Reframing Incidental Exposure’, Journalism 21(8), 1067–82.
- Toff, B., Badrinathan, S., Mont’Alverne, C., Ross Arguedas, A., Fletcher, R., Nielsen, R. K. 2020. What We Think We Know and What We Want to Know: Perspectives on Trust in News in a Changing World. Oxford: Reuters Institute for the Study of Journalism.
- Toff, B., Badrinathan, S., Mont’Alverne, C., Ross Arguedas, A., Fletcher, R., Nielsen, R. K. 2021. Listening to What Trust in News Means to Users: Qualitative Evidence from Four Countries. Oxford: Reuters Institute for the Study of Journalism.
- Tsfati, Y., Cappella, J. N. 2003. ‘Do People Watch What They Do Not Trust?: Exploring the Association between News Media Skepticism and Exposure’, Communication Research 30(5), 504–29.
- Van der Wurff, R., Schönbach, K. 2014. ‘Audience Expectations of Media Accountability in the Netherlands’, Journalism Studies 15(2), 121–37.
- Wenzel, A. 2020. Community-Centered Journalism: Engaging People, Exploring Solutions, and Building Trust. Champaign, IL: University of Illinois Press.
1 To ensure variation in our measurement of trust, we also included some brands that cater to more niche audiences along ideological lines.
2 When we tested the forced-response question using a sample in the UK and an alternatively worded question that included a middle category of ‘neither trust nor do not trust’, we found 28% of respondents selected the midpoint, considerably reducing the percentage in the ‘trust somewhat’ category.
3 When we examine what specific platforms respondents said they had used in the previous week, we find that WhatsApp was used nearly universally in Brazil and India, but by only 57% of people in the UK, and by just a fraction in the US (16%). About two-thirds in Brazil and India also said they got news from WhatsApp compared to just 1 in 10 elsewhere. YouTube, Google, and Facebook were also more widely used in Brazil and India – both as a source of news and in general. In India, we also asked about the use of Telegram and Signal. Half of the sample there (50%) said they had used Telegram for any purpose and 28% said they had done so and received news that way. A smaller segment (12%) said they had used Signal, including 8% for news.
4 Relying on social media for news more generally was not related to trust in news. On average, we find similar levels of trust towards the 15 news brands we asked about separately, regardless of whether respondents said they themselves generally got news from digital or conventional sources.
5 The one brand in the US with a smaller partisan gap was Breitbart, although these differences are larger if the share of respondents who said they had never heard of the brand (40%) are excluded from calculations.
6 In the US, where partisan identification was measured using a two-part question, we include independents who say they lean towards one party or the other.
8 There is just one institution that is an exception to this dynamic. In Brazil, the generally untrusting of news brands are somewhat more trusting of President Bolsonaro’s national government (52% say they completely or somewhat trust it), compared to 42% of those who are generally trusting of news brands. That said, only 40% of those who trust no brands at all say they trust the national government completely or somewhat.
9 In India, although older respondents are overrepresented among the generally untrusting, people under 35 are not necessarily more likely to be generally trusting. Instead, they make up 46% of the untrusting group and 39% of the trusting group. People under 35 in India are more concentrated among the selectively trusting.
10 In India, 28% of the untrusting group did not have a college degree versus 12% of the trusting group. In the UK, those without college degrees were 74% of the untrusting group and 64% of the generally trusting. In the US, the gap was larger: 73% versus 58%.
11 In the US, the differences are not statistically significant for those in the lowest income brackets.
12 When income and trust are treated as continuous variables, we found a significant positive linear relationship between the two in India, the UK, and the US, and a significant negative relationship in Brazil.
13 In Brazil and the UK, people who are not affiliated with any religion are 19% and 42% of the untrusting, respectively, compared to 10% and 27% of the generally trusting, respectively.
14 In Brazil, white people make up 45% of the untrusting versus 32% of the generally trusting.
15 Although the survey was fielded after President Joe Biden was inaugurated, we included the former president in the questionnaire because we assumed that Trump likely remained the more polarising figure in American politics.
16 In Brazil, urbanity was classified based on the population of the city or town where respondents lived. This method may fail to account for qualitative differences between living in an urban or non-urban area. In India, the UK, and the US, respondents selected a response category they felt best characterised their community.
17 We exclude Brazil from this portion of the analysis because responses to this particular question were overwhelmingly one-sided, which may be due to cultural differences, translation issues, or social desirability differences induced by the survey mode.
18 This question was first proposed in 1948 by political scientist Elisabeth Noelle-Neumann (see Delhey, Newton, and Welzel 2011).
19 In India, 30% of respondents who say they trust people from other parts of the country are in the generally trusting group, compared to only 9% among those who say they do not trust people from other parts of India. In the UK, 31% of those who say they trust people of a different race or ethnicity are generally trusting of news versus only 14% among those who do not trust people of a different race or ethnicity.
20 People also tend to report feeling less understood by news organisations depending on where they live, but only in some of the countries studied. People in the UK and in the US (where the relationship between urbanity and trust is also the strongest) tend to feel less understood by news organisations the more rural the place where they live. For example, in the US, 70% of people living on farms or in country villages believe news organisations fail to understand people like them, compared to only 34% of people living in big cities. We see a similar pattern in the UK: around half of the people living on farms, in towns, or in small cities think news organisations don’t understand them, compared to 42% of people in suburbs and 36% of people in big cities. The relationship between the two is much weaker in India and non-existent in Brazil.
21 In Brazil, just 1% of the sample responded in this way – perhaps a consequence of the survey being administered by telephone there instead of online.
22 Lower percentages among the generally untrusting may also be driven by higher susceptibility to social desirability biases on this question.
23 Another possible factor for why results in Brazil are generally lower may be due to the Brazil survey being conducted over the telephone rather than online, which may have elicited somewhat different response styles or assertions of confidence.
24 In the UK, 62% of those who interacted with a journalist in any of the ways we asked about are also very or completely familiar with the difference between an editorial and a news story. This compares to just 28% of those who never interacted with a journalist who say the same. In India and the US, the percentages are, respectively, 69% and 63% for those who interacted with journalists and 55% and 43% for those who have never interacted with journalists. Considering this is one of the concepts fewer people report to be familiar with, the results reinforce a possible sophistication among audiences who interact with journalists.
25 There is an exception to this pattern when we divide those who said they trusted all 15 brands in Brazil from the rest of the generally trusting segments of the public. While in India, the UK, and the US, those who trust all brands are somewhat more familiar with the terms, in Brazil, that dynamic does not appear. Just 45% of those who trust all brands say they are familiar with the difference between a reporter and a commentator, for example, less than the 51% and 54% of the selectively and generally trusting respondents there, respectively. It is possible that in Brazil, people who report trusting all brands may do so somewhat more naively or at least in ways that do not as directly align with self-reported familiarity with how journalism works.
26 We do find one exception to this pattern in India, where generally trusting respondents sometimes have more negative views about journalists than the generally untrusting: 44% of Indian respondents who generally trust news brands say they think journalists get paid by their sources very often, while 31% of those who are generally untrusting say the same. This also occurs when we ask how often respondents think journalists intentionally provoke to get attention for themselves: 37% of those who are generally untrusting say journalists do so very often compared to 42% of those who are generally trusting. We find the opposite pattern in Brazil, the UK, and the US.
About the Authors ↑
Dr Benjamin Toff leads the Trust in News Project as a Senior Research Fellow at the Reuters Institute and is an Assistant Professor at the Hubbard School of Journalism and Mass Communication at the University of Minnesota. He received his PhD in Political Science from the University of Wisconsin-Madison and a bachelor’s degree in Social Studies from Harvard University. He was a Postdoctoral Research Fellow at the Reuters Institute from 2016 to 2017. Prior to his academic career, Dr Toff worked as a professional journalist, mostly as a researcher at the New York Times from 2005 to 2011.
Dr Sumitra Badrinathan is a Postdoctoral Research Fellow who works on the Trust in News Project. She holds a PhD in Political Science from the University of Pennsylvania. Her research interests include studying misinformation, media effects, and political behaviour in India using experimental and survey methods.
Dr Camila Mont’Alverne is a Postdoctoral Research Fellow who works on the Trust in News Project. She holds a PhD in Political Science from the Federal University of Paraná, Brazil. Her main research interests are in the area of political communication, focusing on political journalism, media trust, and media and elections.
Dr Amy Ross Arguedas is a Postdoctoral Research Fellow who works on the Trust in News Project. She obtained her PhD from the Department of Communication Studies at Northwestern University in 2020. Before pursuing her doctorate, Amy worked as a journalist for five years at the Costa Rican newspaper La Nación.
Dr Richard Fletcher is a Senior Research Fellow and leads the research team at the Reuters Institute. He is a principal investigator of the Trust in News Project. He is primarily interested in global trends in digital news consumption, comparative media research, the use of social media by journalists and news organisations, and, more broadly, the relationship between technology and journalism.
Professor Rasmus Kleis Nielsen is the Director of the Reuters Institute and Professor of Political Communication at the University of Oxford. He is a principal investigator of the Trust in News Project. He was previously Director of Research at the Reuters Institute. His work focuses on changes in the news media, political communication, and the role of digital technologies in both. He has done extensive research on journalism, American politics, and various forms of activism, and a significant amount of comparative work in Western Europe and beyond.
For their assistance in conducting this research, the authors thank Jean Estevão de Souza, Luciana Chong Rodrigues, and Marlene Treuk from Datafolha; Ahmar Kamal and Vijay Kumar from Internet Research Bureau; and Martha Espley, Maki Hasegawa, Julie Soulsby, and Trevor Vagg from Kantar. We are also grateful to Kate Hanneford-Smith, Alex Reid, Rebecca Edwards, and Louise Allcock for helping to move this project forward and keeping us on track, along with the rest of the research team at the RISJ for their feedback and input on this manuscript.
Published by the Reuters Institute for the Study of Journalism with the support of the Facebook Journalism Project.