Ipsos MORI Thinks

Social trust: how much do we trust each other?

RACHEL ORMSTON Research Director

Social trust – the trust we have in each other – has preoccupied researchers and commentators for decades. While the impact of distrust in politicians on our inclination to vote in elections may be obvious, the consequences of lack of trust in our fellow citizens are arguably far more damaging and wide-ranging. Social trust has been linked not only with democratic engagement, but also with national economic growth, quality of life, and even rates of suicide.34 Being able to trust each other is central to Putnam’s concept of social capital – when people get together, their trust in each other and their ability to work collectively to solve social problems is increased.35 In this context, anxiety about social trust is about much more than just how we feel about each other – but about the wider ripples low social trust creates for the democratic, social and economic wellbeing of society.

As David Halpern argues:

“In short, it’s not much fun living in a place where you don’t think other people can be trusted. Low trust implies a society where you have to keep an eye over your shoulder; where deals need lawyers instead of handshakes; where you don’t see the point of paying your tax or recycling your rubbish (since you doubt that your neighbour will do so); and where you employ your cousin or brother-in-law to work for you rather than a stranger who would probably be much better.”36

So just how much trust do we have in each other? And do we trust each other less than we used to? In this article, we take an overview of trends in survey data around the world to try to answer this question.

The most common measure of social trust is based on some variation of the question:
“Generally speaking, would you say that most people can be trusted or that you can’t be too careful in dealing with people?”

A variant of this question has been included regularly on the World Values Survey and European Values Survey (since 1981), Latinobarometro (since 1996), European Social Survey (since 2002), the General Household Survey in the USA (since 1972) and British Social Attitudes (since 1998), so we can compare patterns across countries and over time.

Scandanavia tops the trust tables

While the exact figures produced by different surveys vary, in all the surveys in which they feature, Norway, Denmark and Sweden feature in the top five most trusting, with 60%+ saying they feel most people can be trusted. The Netherlands, Finland and Iceland also occupy positions near the top (for example see the most recent wave of ESS), as does China (60%) in the most recent World Values Survey (WVS) (2010-14).

At the other end of the trust rankings are many Latin American countries. In the most recent Latinobarometro (2017) only Ecuador recorded trust levels above 20%. The small number of African countries included in the WVS tend to have low levels of social trust as well. Eastern European countries also tend to be less trusting than those in the west of the continent, with some notable exceptions – France, for example, falls in the bottom half of the social trust rankings in both the most recent European Social Survey (ESS) and the European Values Survey (EVS). Meanwhile, trust levels across Asia vary widely, from 9% in Malaysia to over 60% in China (WVS, 2010-14).

Declining social trust is not a uniform global phenomenon

Examining patterns over time also shows that, in spite of media narratives to the contrary, social trust is not declining in a uniform manner across the world – there are clear variations in trends in trust between countries.

Americans have become more inclined to bowl alone

In the USA – the focus of Putnam’s seminal work on declining social capital – there is indeed some evidence for a long-term fall in social trust, although much of the decline occurred by the mid-1990s. The General Social Survey shows that, in the 1970s and 1980s, the proportion of Americans who said most people could be trusted fluctuated around the 40% mark, before falling to the mid-30s by the early 1990s, and to the low-30s from the mid to late 2000s. The World Values Survey also shows declining trust in the United States, falling from 44% in 1981-84 to 35% by 1995-98 (and staying at this level in the most recent data, from 2010-14).

However, while social trust remains at a historically low level, there is no evidence that the apparent sharpening of political divides in the USA since the election of Donald Trump in 2016 has led to any

further decline. The 2018 GSS figure puts trust at 31% – on a par with figures from 2006 onwards. However, if general trust in others has not changed since Trump, our willingness to trust people whose political beliefs differ from ours may have suffered more. Research has highlighted a steep increase in political polarisation in the US – the proportion of both Democrats and Republicans who hold ‘very unfavourable’ views of their counterparts in the other party has rocketed since the mid-1990s.37 So while Americans’ general propensity to trust each other has not fallen any further over the last decade, the intensity of distrust between partisans may well have increased.

Long-term downward trend in Latin America (but a lot of movement in between)

Only two of the 18 countries included in the most recent Latinobarometro (Ecuador and Chile) recorded higher levels of social trust compared with the earliest year of the survey in 1996. Across the other 16 countries, people’s trust in their fellow citizens has either stayed at its previous low level (Costa Rica, Peru) or fallen, by 14 percentage points or more in the cases of Guatemala, Panama, Paraguay and Uruguay.

However, the decline in social trust apparent across Latin American countries has not been linear. In Venezuela, for example, trust increased from 11% in 1996 to 25% in 2007, before falling back again from 2013 onwards, following the death of Hugo Chavez and the subsequent political and economic turmoil, to its current low of 9%. Other Latin American countries have also seen similar rises and falls over the last 20 years.

Trends in trust vary widely across Europe (and between surveys)

Establishing a clear picture of trends in trust across Europe is complicated by differences in trends between different studies (WVS, EVS and ESS). For example, the WVS finds that trust has fallen since the 1990s in six out of the eight Eastern European countries included in the 2010-2014 wave. The ESS, in contrast, shows mean trust scores increasing between 2002-06 and 2012-16 for four of these countries!

However, what is clear is that there is no evidence of a Europe-wide decline in social trust. Indeed, a number of countries have seen their levels of trust in other people increase substantially in recent years. Scandanavia (Sweden, Norway and Denmark) has become even more trusting, improving on an already high base, while Germany and the Netherlands also recorded double-digit increases in social trust from the 1981-84 to 2008-10 waves of the EVS.

Meanwhile, in Britain, recently published British Social Attitudes data indicates that, after remaining broadly constant from 1998 to 2014 (at 45-47%), the proportion who say most people can be trusted actually increased to 54% between 2014 and 2017. Very similar patterns of rising trust in the ordinary man or woman in the street are apparent from Ipsos MORI’s annual Veracity Index. So in spite of the apparently extreme divides in Brexit Britain, if anything, Brits are becoming more inclined to trust each other, not less.

There is no clear trend across the rest of the world

Levels of social trust have also gone up as well as down elsewhere in the world. In Australia, for example, social trust fell back between the 1981-84 and 1995-98 rounds of the WVS, before increasing again in 2010-14. In India, trust fell from 34% in the early 1990s to just 17% in 2010-14 (WVS). Trust has also fallen in recent years (to a lesser extent) in South Korea and Taiwan, but has held steady at around 60% in China. In short, while a number of countries have  seen social trust fall in recent decades, there is no single consistent pattern – claims of a ‘global crisis’ are overstated.

What explains diversity in social trust?

Theories about what underpins social trust abound. Recent work for the IMF has explored the impacts of economic inequality on social trust, arguing that inequality lowers people’s sense of fairness and therefore their trust in others (Gould and Hijzen, 2016). This may help explain why the USA, where income inequality has increased over time, has seen declining trust, while high levels of trust have held steady or improved in countries like Sweden and Denmark where income is more equally distributed. Other research has looked at demographic and generational patterns to trust – for example, Ipsos MORI’s ‘Generations’ project in the UK found that while overall, most people become more trusting with age, there are also marked cohort effects. Millennials (those born between 1980 and 1995) were markedly less trusting as young adults than previous generations, though they have become more trusting with age. In contrast, however, ‘Generation Z’ (those born 1996 onwards) are much more likely to say they “generally trust the man or woman in the street to tell the truth” than their suspicious predecessors at the same age. In the US, generational divisions in trust are even more apparent. As Baby Boomers have entered older age, they have shown a pronounced decline in trust, while Generation X and Millennials started off with lower social trust and have remained less trusting.38 The relationship between social networks – who you spend your time with – and social trust is another ongoing theme (e.g. Li et al, 2019). In short, evidence suggests that the number of people in your social circle and their social status help predict how much you trust other people in general.

As overall patterns and trends in trust vary substantially between different countries, it is also likely that the drivers of trust vary across the globe. What is clear is that claims that people across the globe are becoming less likely to trust each other are premature. That is not to say that there is no reason to be concerned. Even in countries where trust is not continuing to decline, it is clear that the bar is set fairly low – with some notable exceptions, a majority of people across most countries are unwilling to trust their fellow citizens.

This is a major problem. ‘Transactional costs’ will be higher for individuals and companies (as we formalise our exchanges because we don’t trust each other to honour informal agreements). We won’t work effectively to solve collective problems – particularly when a lack of trust is combined with antipathy to people of a different political persuasion to our own. And our quality of life and even national economic growth may be hampered.

There is clearly no easy answer to restoring our trust in each other – if there were, governments would have done so decades ago. Action needs to focus on both the macro- and micro-levels. At a macro-level, given evidence that the degree of economic inequality within countries is a strong predictor of levels of social trust, particularly in advanced economies,39 addressing the unequal distribution of wealth may also help to boost trust. At a micro-level, creating more opportunities for us all to mix with people who are not like us – including our political ‘opposites’ – to tackle increased social and political polarisation could, in turn, help us re-build our trust in each other. More generally, we need to look to those parts of the world where trust is high and growing for lessons on how to support and build trust between citizens, so that more of us can enjoy the economic, social and cultural benefits of a ‘high trust’ society.

“The ‘decline in trust’ has not been linear. In Venezuela, for example, trust increased from 11% in 1996 to 25% in 2007, before falling back again from 2013 onwards, to its current low of 9%V”

Trust: The Truth?

We decided to write this report because we wanted to test if the prevailing narrative matched the data. The ‘truth about trust’ is that trust is complex, and takes many forms (many of these forms are not in crisis or decline). Without some degree of trust society simply would not function…