In addition to the survey research described throughout this report, we have also explored trust through eight focus groups in four countries: Spain, Germany, the US, and the UK. Taken together, the focus groups generated a number of specific hypotheses that we were able to put to the test using the survey data.
This subsection will primarily draw on data from six countries. We look in detail at the four countries for which we have focus group data, as well as Finland and Greece: two countries that are at opposite ends of scale in terms of overall trust in the news.
PROPORTION THAT AGREED THAT THEY CAN TRUST MOST NEWS MOST OF THE TIME
|Strongly agree||Tend to agree||Neither||Tend to disagree||Strongly disagree|
Base: Total sample in each country.
Trust in the news does not map particularly well onto demographic variables. When we look at different countries, we find that different groups are more likely to trust the news. However, almost everywhere younger people tend to trust the news slightly less than older people, even after controlling for variables such as gender, income, education, and politics. 1 In Germany, for example, over half (55%) of over-35s say they trust the news, but this figure drops to 41% among under-35s. The difference in the US was too small to be significant.
TRUST IN THE NEWS AMONG UNDER-35s AND OVER-35s
|Under 35s||Over 35s|
Base: Under-/Over-35s: Finland = 556/1485, Germany = 484/1551, UK = 426/1598, Spain = 639/1465, US = 504/1693, Greece = 598/1438.
Political beliefs are also linked to trust in the news. However, the picture is complicated. In countries where trust is high, such as Finland, political identification tends not to have an impact on trust, with each group equally likely to trust the news. However, in the UK, although overall trust is fairly high (51%), right-wingers are more likely to trust the news than those on the left.
TRUST IN THE NEWS AND POLITICAL IDENTIFICATION
Q1F. Some people talk about ‘left’, ‘right’ and ‘centre’ to describe parties and politicians. With this in mind, where would you place yourself on the following scale?
Base: Left/Centre/Right: Finland = 389/824/355, Germany = 149/1485/69, UK = 468/1009/292, Spain = 626/1130/159, US = 476/871/591, Greece = 355/1231/173.
Note: Respondents that selected “Don’t know” at Q1F were removed from the analysis.
A consistent theme from the focus group sessions in all four countries was the view that trust in the news is strongly tied to trust in specific news brands.
We were able to put this to the test using our survey data. When we look at the strength of the association (or correlation) between overall trust in the news and a number of other potential drivers, we can see that there is a very strong link between trust in the news and trust in news organisations (0 indicates no association, and 1 indicates exact overlap). In all 26 countries trust in news organisations is the most important driver of overall trust, and is significantly more important than trust in journalists, and freedom from undue governmental influence. In other words, trust in the news is almost synonymous with trust in news brands. Of the different dimensions examined, perceived freedom from commercial influence was the least important driver of trust in most countries.
Scroll data area to see more
DRIVERS OF OVERALL TRUST IN THE NEWS
|Trust in news organisations||.74||.82||.77||.76||.83||.73|
|Trust in journalists||.68||.73||.64||.65||.75||.62|
|Freedom from politics||.49||.65||.53||.48||.61||.46|
|Freedom from commerce||.49||.61||.50||.50||.58||.42|
Base: Total sample in each country.
Note: A Pearson’s product moment correlation test was used to compute the correlation coefficients. A Steiger’s Z-test was used to test for differences between them.
Another point often made by our focus group participants was that trust in news brands takes a long time to build. Some news brands – typically those that have been around a long time – are often seen as main sources of news, whereas new players – even if they have a large reach – are thought of as secondary sources or ‘guilty pleasures’.
If we look at the proportion of users of each digital-born news brand who say that this is their main online news source (e.g. the percentage of users of Yahoo News who say it is their main news brand), we can see that the figures tend to increase for brands that have been active for longer. Indeed, if we group together all 32 of the digital-born news brands included in our survey in Germany, Spain, the UK, and the US, we see a strong correlation (.65) between the age of the brand and the proportion who say it is their main online news source. Regardless of their overall reach, brands like Yahoo, MSN, and AOL, which are now over 20 years old, are more likely to be a main source of news than brands like BuzzFeed and Huffington Post, which emerged in the last decade. New players like Vox in the US and El Español in Spain may take encouragement from the fact that their prominence in the minds of consumers will likely increase slowly over time. The main exception to this trend is Vice which, despite being present in one form or another since the mid-1990s, is only seen as their main source of news by 5% of its users in the UK and 8% in the US.
PROPORTION OF USERS OF DIGITAL BORN BRANDS THAT SAID EACH WAS THEIR MAIN BRAND, ORDERED BY AGE OF BRAND (IN BRACKETS)
Base: All who used The Lad Bible/BuzzFeed/Huffington Post/MSN News/Yahoo News Vox/BuzzFeed/Huffington Post/MSN News/AOL News in the last week: UK
= 71/183/337/145/151, US = 78/340/562/369/150.
Base: All who used BuzzFeed/Huffington Post/Gmx.de/t-online/Web.de ElEspanol.com/Eldiaro.es/Publico.es/Huffington Post/ElConfidencial.com: Germany = 32/161/267/298/300, Spain = 148/372/290/302/421.
Note: The age of each brand was taken from Wikipedia. A Pearson’s product moment correlation was applied to the data from UK, US, Germany and Spain. The test showed that there is a positive linear relationship between age of brand and proportion of users that said it was their main source: r(31) = .65, p < 0.01.
- A series of ordinal regression models were used to test for significant associations with trust as measured on a five-point scale, and to simultaneously control for the influence of other demographic variables. ↩