Is trust in crisis, or is it just complicated?
GIDEON SKINNER Research Director
You don’t have to look very hard to see media stories about trust being in crisis. This obsession with ‘trust’ among politicians, business leaders and the media is in part triggered by recent so-called ‘populist’ revolts against elites around the world. But when you look at the data in detail, it paints a much more complex picture that suggests sweeping statements like this need to be taken with a pinch of salt – especially when you combine it with a more sophisticated understanding of what ’trust’ really means, as discussed elsewhere in this report.
The purpose of this report is to take a look at what the survey data on public attitudes to trust from around the world actually shows. For a topic as huge as trust, that has so many implications across the personal, public and private spheres, we are not going to pretend that it can be reduced to a single simple soundbite, nor that what happens in one country or for one type of trust holds true for all others.Instead, we believe that the first stage of understanding what is happening with trust, and how we might improve it in the future, starts with an open-minded look at what the evidence really shows, rather than with some of the myths that have developed over the years. The rest of this report goes into this evidence in more detail, but what are some of the overall conclusions?
Trust is low in many arenas, and often has been for years. Low levels of trust seem to be a chronic issue, rather than a very new acute one.
Looking at the absolute levels of trust around the world shows there is little room for complacency. It is rare that the majority of the population in any country thinks that most of their fellow citizens can be trusted, and in many countries, it is significantly less than that. There are also low levels of trust in many of the key institutions that society relies on – especially those of government and politics, the media, big business and more.
At the same time, low levels of trust are not new. The respected British Social Attitudes Survey (BSA) shows that trust in others in the UK has changed little over the last twenty years, and going even further back, the General Social Survey (GSS) in the United States found even in the early-1970s only a minority felt that most people could be trusted – and it has been falling steadily since then, not just recently. Data from Pew suggests that a majority of Americans said they trusted the government in Washington most of the time in the 1950s and 1960s, but this had disappeared by the mid-1970s. Even during the Second World War, a famous Gallup study found that only a minority of Britons felt that politicians were doing their best for the country (though it has dropped even further since then). Furthermore, as we discuss in more detail below, there are wide variations in levels of trust between different countries, different institutions and different groups of people.This suggests that, on the one hand, talk of a worldwide crisis in trust as a new phenomenon is misplaced; low levels of trust are not new, and indeed there are plenty of examples of rising trust that can be found. But on the other hand, just because it is not new, it does not mean there is nothing to be worried about. To take the analogy further, if trust is a chronic issue rather than an acute one, it implies that it has been developing over the long term, and may have multiple causes rather than a single, simple diagnosis. It might also reduce society’s resilience against other sicknesses.
Very few people think trust has been rising over the last twenty years
Nevertheless, the perception of large numbers of people around the world is that trust has been falling over the last twenty years or so, as shown in the chart overleaf. Perceptions that trust is in decline are particularly high for the institutions we see with some of the lowest levels of absolute trust – the government and the press – but even for our trust in others, slightly more think it has fallen than stayed the same, and very few think trust has actually risen.In some cases this will simply reflect reality, but in other countries where survey data suggests trust is stable (or ev.0en sometimes showing signs of improvement), this may reflect our negativity bias that we see in much of our other perceptions work. It can be both fed by, and feed into, media stories of trust in crisis. For example, around four in ten in the US believe that levels of trust in others have fallen in the last twenty years, which reflects a real (though small) fall in the trends there over that time. But four in ten Swedes also believe trust has fallen – when in fact it has stayed at a relatively high level over that time, as do four in ten in Brazil (where trust has stayed at the same low level over that period). And four in ten Brits also believe that trust in others has been falling – where if anything the trend is slightly up. There have also been rises in trust in others in Germany and Belgium since the early 2000s, but fewer than one in ten in those countries think that trust is rising.
Perceptions that trust in governments is falling are even stronger. Again, in several countries this reflects what survey data is showing – between around six and seven in ten in the US, France, Spain, Italy, Chile and Brazil all believe that fewer people trust the government than did twenty years ago, all of which is in line with trends in survey data in those countries.
But people in countries that have seen little change in trust in government, or even improvements in some European countries, don’t seem to recognise this. So again, almost just as many (around six in ten) in Sweden and Germany believe that trust in government is falling – but in those countries, measures of trust in government are, if anything, higher now than they were roughly twenty years ago.2
A crisis in confidence?
Much of the data looking at trust in key institutions is phrased as “confidence”, itself instructive in showing how people frame this issue. In the US, according to the General Social Survey (GSS), confidence in a whole range of institutions is at record lows. This applies to institutions of government (confidence in Congress fell from 24% in 1973 to 6% in 2018; in the federal executive from 29% to 12%), private sector companies (confidence in banks fell from 32% in 1975 to 19% in 2018; in major companies from 31% in 1973 to 20% in 2018), and other parts of the public sphere (confidence in education fell from 37% in 1973 to 25% in 2018; in organised religion from 36% to 21%).
Again, though, there are signs that this is a long-standing, chronic issue – confidence was never very high in any of them, and has been falling since the 1970s, not just in the last few years – and some institutions, such as the military, buck the trend. Perhaps just as importantly, it is not clear in which direction the causality lies. Is there a crisis of confidence in institutions because trust is falling, or is trust falling because institutions are not seen to be delivering for their citizens? This would reflect the strong ‘system is broken’ sentiment that we see at Ipsos through our global polling. Even if trust is falling, which particular aspect of trust is to blame? Simply bemoaning low levels of such a large and multifaceted issue as trust does not pinpoint what actions to take to improve it.
The trends in confidence in European countries are less clear cut and show a similarly mixed picture. According to the European Values Study (EVS), between the 1981/84 and 2008/10 waves, on average confidence fell for churches, but for many other institutions, such as parliament, the press, and major companies, confidence was little changed over those twenty years, and for several others (such as the army, education, civil service and labour unions) confidence actually increased (although this hides variation by country). The latest results of the EVS are yet to be published, which will tell us whether confidence in civil institutions in Europe has been affected by the last decade or so of austerity, but evidence from other surveys – such as the European Social Survey and Eurobarometer – suggests that the trend in Europe at least is slightly less negative than elsewhere.
There is some limited evidence that confidence is improving in the last couple of years. Ipsos measured public confidence in seven institutions in 21 countries around the world in 2016 and 2018: international institutions, governments, banks, the media, big companies, the justice system and political parties. On average, confidence remains low for all of them – especially the media, government, and political parties – and had only changed marginally over the two years. However, that does hide a small number of more positive country changes – notably in Spain, Germany and Mexico. We can also detect surges in confidence in government with new leaders such as Trudeau, Modi and Macron – or at least when they were new.
This brings us to one of the clearest findings when looking at the data on public opinion on trust: you can’t take a one-size-fits-all approach to different countries around the world. There are differences in question wording, approach, and fieldwork dates, so we have to caveat the comparisons to some extent. But nevertheless, the trends look very different – and not just on the confidence questions quoted previously. Take social trust (i.e. trust in other people), for example. It’s not just the US where social trust looks to have fallen – there are also falls in many Latin American countries, in India and some African countries too (taking an overview of data from the World Values Survey and Latinobarometer). But there are also countries where trust seems to be rising, and not just in Europe – rises have been recorded in Australia and New Zealand, and even a small rise in China (again according to the World Values Survey).
Even within Europe itself, there are clear differences by country. Since the turn of the millennium, according to Eurobarometer and the European Social Survey, there are countries that have seen a rise in trust in institutions, and those that have seen a fall. The countries where trust is on the up tend to be richer and in the north and centre of Europe – such as Germany, Austria, Switzerland, the Netherlands, Norway and Sweden, though also including Poland and the Czech Republic. Falls in trust on the other hand are particularly noticeable in the Mediterranean countries of Spain, France and Italy, and also Romania and Slovenia.
Trust in what?
Nor is it the case that all institutions are suffering from the same levels of (mis)trust. As the Ipsos MORI Veracity Index makes clear, the public distinguishes between different professions, and a similar pattern is found when comparing trust or confidence in institutions, from surveys such as the GSS in the US, and Eurobarometer, the European Social Survey and the European Values Study in Europe. Public services, such as education, healthcare/medicine, the police and the armed forces tend to receive more positive trust ratings from the public. Institutions such as banks and major companies, government, the press and parliament all receive low scores, but the very lowest levels of trust seem to be reserved for politicians and political parties. That reflects the low opinion many people living in democracies have for the people they vote for – but again, not a finding that is particularly new or surprising. Of course, distrust of politicians does not necessarily mean we prefer autocracies or direct democracy. In the UK, every time we have asked the public if they want to replace elected politicians with professional managers the answer has been no!
Indeed, there are also examples of institutions where trust has actually been rising over time. As has already been noted, according to the GSS, confidence in the military rose from 33% in 1973 to 53% in 2016, and it rose from 56% to 62% on average in countries taking part in the European Values Study in 1980/1 to 2008/10. And there are more examples on a country-by-country basis. Even in the US, confidence levels in organised labour and the scientific community are little changed from the 1970s. In Europe, according to the European Social Survey, trust in the police rose in 14 countries between 2002 and 2016, and trust in the legal system in twelve. There are even a small number of countries, including Germany, with rises in trust in political parties, albeit from very low bases.
The press and the internet
One aspect of our lives that is new, of course, is the internet and social media. Its rise and changing public opinion towards it may help to explain why there is such a strong feeling that trust is in crisis, even if that isn’t always borne out by the data as a whole. One useful tracker of trust in the internet comes from Eurobarometer, which has been measuring it since 2006. On average across the 29 countries included from 2006 to 2018, trust in the internet was little changed, from 38% to 36%. However, in 2006 the proportion answering “don’t know” stood at 31% – but twelve years later that figure had halved to 15%, while those not trusting the internet rose from 31% to 49%. The more Europeans saw of the internet, the more they decided they did not trust it.
It is useful to compare this with trust in the press in the same 29 countries over the same period. Unlike trust in the internet, trust in the press was little changed on average over those twelve years (although with a dip in the first six years, which then recovered), and is higher than trust in the internet, with 48% saying they trusted the press in 2018 and 44% not. The difference is starker at a country-by-country level. Twenty-two of the 29 countries measured saw a fall in trust in the internet over that time. However, of those 22, half actually saw a rise in trust in the press over the same period.
The same comparisons are not available for other countries, and not all the long-term trends for trust in the press elsewhere are as positive. For example, according to the General Social Survey in America, confidence in the press and in television in the US is lower in the 2000s than it was in previous decades, and the World Values Survey has also seen long-term falls in confidence in the press in other countries – such as Australia, Brazil, Mexico and Russia – although rises in other places such as India, Japan and South Africa.
Again, in Europe the picture is slightly different, with the European Values Study showing on average little change in confidence in the press between the 1980s/90s and 2008/10, with as many countries showing rises as falls. Nevertheless, there is more evidence to show that the internet, and social media, remains significantly less trusted as a source of information than more traditional media. According to a 2018 Eurobarometer study, over six in ten say they tend to trust the news they receive from the radio, television and printed news. But only half trusted online news, and just a quarter news from social media.
A 2016 Pew study3in the US found that trust in information from national and local news organisations is about twice that of trust in information from social media, while research from Reuters4 showed that in five of seven countries, trust in digital media was lower than public service broadcasters, commercial TV or print.
There seems to be a bit of a paradox here. On the one hand trust in some traditional news outlets, and indeed in fellow citizens, seems to be holding up despite low trust in the internet and social media – which will of course involve interacting with the very same people and same media companies. This may suggest a welcome lack of naivety in our interactions online, as well as an equally welcome ability to distinguish between the sound and fury on social media and our dealings in the rest of our lives. But on the other hand, given the growth in internet usage, and the extent to which it is becoming out-of-date to talk about a distinction between our online and offline lives, these low levels of trust should be a concern – particularly given other trends we are beginning to see of echo chambers and polarisation.
Beneath the surface – growing polarisation?
One of the drivers of the ’Age of Uncertainty’ that Ipsos has described in the United States is a growth in tribalisation, and specifically a negative type of polarisation where people do not just stand up for their own views, but are increasingly antagonistic towards those with alternative views. This started before the modern age of social media – GSS5 data shows Democrats and Republicans becoming more polarised over abortion since the mid-1980s, for example – but it continues to get worse. According to Pew,6 between 1994 and 2016 the proportion of Republicans with a very unfavourable view of the Democratic Party rose from 21% to 58%, while the proportion of Democrats with a very unfavourable view of the Republican Party rose from 17% to 55%.
There are also some signs of polarisation in other countries, although the evidence is not as strong as in the US (and it is also less clear whether this is a new phenomenon or a long-term issue). Analysis by Schwarz and Draca7 using the World Values Survey has found the US stands out for its increase in polarisation among citizens, but nevertheless still finds ‘muted’ evidence for increasing polarisation and a ‘disappearing centre’ in some other countries. The US also stands out as having the most polarised online media in a 2017 Reuters Institute analysis,8 where there is also evidence of polarisation in southern and central European countries such as Italy, Spain and Poland.
In the UK, analysis by Jonathan Wheatley of Oxford Brooks University9 has observed a growing gap between supporters of parties of the left (Labour, the Greens and the LibDems) and parties of the right (the Conservatives and UKIP) between the 2015 and 2017 general elections. The gap is not just there on issues of left- and right-wing approaches to the economy, but is growing even more on cultural/open vs closed values (such as attitudes to immigration, foreign aid and LGBT issues).
The Brexit vote has put even more emphasis on the cultural divide that it revealed, along with data showing that whether you voted ’remain’ or ’leave’ has become more important to people’s political identity than traditional party loyalties, and this, if anything, has increased since the referendum.10 Other research has suggested it is the Remain side that is particularly insular – for example 37% of Remain voters said they would mind if one of the family married a Leave voter, but only 9% of Leavers felt the same of Remainers.11 Our own analysis of Ipsos MORI’s satisfaction trends with party leaders over the last 20 years also shows that the gap between Conservative and Labour supporters’ views of their own party leader and those of the opposition leader is growing. This is support for the theory that political polarisation may be growing, although not yet conclusive proof.
Having said that, even though many people believe that their society is becoming more divided, it is still easy to overstate the problem. Ipsos research for the BBC12 shows that around the world, one in three think that political divisions are dangerous for society, but exactly the same proportion think that in fact it is healthy for society to have a range13 of different views. And only three in ten think that those with opposing political views don’t care about people like them. It’s also important not to mix up cause and effect – declining trust could be an outcome of this just as much as it is a driver, and other factors are likely to have an impact: for example, economic insecurity since the Great Recession has been shown to be associated with increases in support for populism and political distrust.14 Nevertheless, feelings of negative partisanship and antagonistic tribalism may be contributing to our perceptions of a crisis in trust. It could be that low levels of trust over the long term have made us more susceptible to polarisation.
We wrote this report because the widespread media focus on trust in crisis made us want to examine what the data actually said about it. As we show, whether there is a new crisis of trust is debatable. But the crash of 2008, the rise of ‘culture wars’ and the deeper drivers of the crisis of the elites that the world is experiencing are driving a focus on trust.
This report seeks to inform that debate on trust, and to highlight what could been done to rebuild it. It will not be simple. We have lived with low levels of trust for decades, but we hope that this report is part of a more realistic appraisal of where we currently are, and what might be done.