Hosking’s book was widely anticipated. I had hoped that it would be a worthy successor to Adam Seligman’s The Problem of Trust.(1) However, it is largely a rambling discourse on concepts that are often barely connected to trust. There is no clear idea of what varieties of trust are. Many of Hosking’s claims are at variance with the evidence we have on trust. And Hosking has little to say about Seligman’s work, other than to mention him critically in passing a few times.

He largely ignores newer work on trust, with only occasional references to Russell Hardin and Robert Putnam, but none to Dietlind Stolle, Bo Rothstein (once cited in passing on p. 2), Kenneth Newton, Sonja Zmerli, Marc Hetherington, Jan van Deth, Mark Warren, Dario Castiglione, Roderick Kramer, Masamichi Sasaki, Per Selle, John Hibbing, Elizabeth Theiss-Morse, Christopher Anderson, Jack Citrin, Barbara Arneil, Toshio Yamagishi, Karen Cook, Tom Tyler, Markus Freitag, Diego Gambetta, or myself. These students of trust (among others) cover a wide range of approaches. They don’t always (or often) agree with each other. But their essays and books frame the current debates over trust.

Beyond Hosking’s failure to engage with other students of trust are other concerns:

1.          I don’t know what Hosking means by trust. He has a discussion stating that ‘… we are all interdependent … Trust is a vital ingredient in this web of interdependence’ (p. 4). He presents a ‘semantic map’ (p. 29) of trust with poles of certainty/ uncertainty and freedom/compulsion. The objects of trust include the reliance of babies on their mothers, reliance on the currency, voters’ consent in a democracy, confidence in buying food in a supermarket, hope (a traveller in an unfamiliar country), loyalty (in marriage), acquiescence (dealing with the police), and interdependence of soldiers in combat. Yet trust is elusive. Hosking does not distinguish among types of trust or the range of trust (or the ‘radius’ as many now call it). 
2.          Trust, Hosking argues, is a feeling, an attitude, and a relationship. Yet trust is more than that. It is not necessarily a relationship, as I argue.(2) Generalized or moralistic trust is not about trusting anyone in particular, nor is it about trusting anyone to do any particular thing (such as repaying a loan). Hosking refers to some discussions of a moral basis for trust and even argues that it is an important component of trust. But this concept vanishes as soon as Hosking introduces it (pp. 36–7).
3.          The contemporary literature has many distinctions about different forms of trust: There is nothing about the distinction between ‘bonding’ (in-group) and ‘bridging’ (out-group) ties.  Nor is there any discussion of the distinction between generalized trust (in people who may be different from yourself) and particularized trust (only in people like yourself). Hosking distinguishes between ‘strong’ and ‘weak’ trust, depending upon the size of the stakes in putting confidence in someone or something, and ‘thick’ and ‘thin’ trust, which reflects the extent of contact between the truster and the trustee. The latter distinction is close to the bridging/bonding and generalized/particularized distinctions. But Hosking makes little of this – and he does not consider at all whether one can trust people you don’t know at all. In the contemporary literature on trust, there is considerable debate over whether there is a ‘transmission belt’ of trust from people you know to strangers. This is one of the central debates in contemporary accounts of trust, but Hosking has nothing to say about it.   
4.          There are extensive discussions of how religion fosters trust, with occasional references to how religion may foster distrust. But there are differences among religions – and within them that Hosking does not consider.
5.          Hosking begins with a detailed chapter on how the Soviet Union destroyed trust (chapter one). The Communist regime was so repressive that Joseph Stalin told Nikita Krushchev: ‘I don’t trust anyone. I don’t even trust myself’ (p. 17).  Distrust in the totalitarian state is supposed to set the tone for the discussion to follow (p. 21), but there is no follow-up until the middle of chapter six, 130 pages later, where Hosking devotes three pages to the familiar story of how post-Soviet Russia still seems very Soviet. He is clearly an expert on Russia and he seems more interested in telling stories about that country than focusing on trust.

The story of low trust in the Soviet Union (and contemporary Russia) is symptomatic of the disorganization of the book overall–and of Hosking’s reliance upon narratives without resort to evidence. His chapters are collections of ‘just so stories,’ narratives that sometimes seem to be removed from the theme of the book.                                                                   

The Foundations of Trust

Hosking’s major claim in the introduction is that we rely upon trust in order to conduct our lives. We trust that the aeroplanes won’t crash and that government officials will perform their duties faithfully, for ‘once generalized social trust declines to a certain point, then it plunges rapidly and damagingly [sic]’. But Hosking never makes a connection between the mechanics of flight and generalized social trust – largely because there is no such linkage. There is also weak evidence at best that bureaucratic performance makes people more or less likely to trust each other. Nor is there any evidence of a downward spiral in social trust. In the United States, generalized trust has fallen from 58 per cent in 1959–60 to about one-third today. But it has remained at this level for almost a decade.  For most countries, the level of trust is rather stable over time  It is resistant to big catastrophes such as aeroplane crashes – or even natural disasters.  When we make such claims, we need to look at real data, not just our speculations or stories.

Trust, Hosking (p. 43) argues, is the default condition for most of our social relations. Yet, cross-national surveys consistently show that in only a handful of countries do a majority of people believe that ‘most people can be trusted’.  The trusting citizens are found in the Nordic countries: Denmark, Sweden, Norway, and Finland generally have the highest shares of trusters. The Dutch, Anglophone Canadians, and (more recently) the Australians also are more trusting. The least trusting countries include Turkey and most countries in Latin America and Africa. What shapes generalized trust? I show that economic equality leads to greater trust across countries, American states, and in the United States over time.(3) This finding is widely accepted in the literature on trust. But the words ‘inequality’ or ‘equality’ appear only once in Hosking’s book, in a quote from Michael Sandel (p. 91) as part of an attack on money and the financial system more generally (see below).

Now much of the difficulty stems from the fact that trust is multifaceted. In a ‘think aloud’ experiment in the 2000 American National Election Study pilot survey, people were asked what the generalized trust question (‘Generally speaking, do you believe that most people can be trusted or can’t you be too careful in dealing with people?’) means. They responded with statements about the goodness (or lack thereof) of others, not on whether aeroplanes crash or even whether people treat them well.(4) There is no evidence that planes are more likely to crash in the United States, where trust is now moderate or in Denmark, where it is the highest in the world.

Nor is there evidence that trust is the default position even in a democratic society. Hosking (p. 177) argues that we can trust our public officials because we can remove them. Yet trust in government is consistently higher in China than in most other countries.  And it has risen and fallen time and again in the United States – with ‘highs’ of 77 per cent during the ‘American High’ of the early 1960s (5) and 54 per cent after the 9/11 attacks. Today it stands at 24 per cent.(6) Trust in government is not the same as trust in people, which has fallen in an almost perfect linear manner in the United States since it was first measured in 1959–60. As my Norwegian friend Per Selle once said, ‘I trust my people, but I don’t trust my government’. To be sure, Norwegians are more trusting of government than are people in many other countries, but even there faith in political institutions goes up and down.

Yes, major disasters such as the tsunami in Japan and the ferry sinking in South Korea can lead to less trust in government. But the linkage is not simple. What mattered in both cases was the perception by a large segment of the public that the disasters were made worse (or happened at all) because of corrupt deals between unregulated companies and the government. Politicians and executives profited at the expense of many helpless victims. In neither case did technical failure lead to a loss of trust. When government responds well – and quickly – to disasters, trust increases. The strongest determinant of confidence in government is the performance of a nation’s economy. Again, it helps to look at the evidence. There are many studies that support the claims I have made.

In God We Trust?

Hosking devotes an entire chapter to ‘Godly Homelands.’ Put simply, religion ‘is the most all-embracing symbolic system’ of trust (p. 50). Regardless of one’s faith, religion binds believers to each other. Hosking’s examples run from ancient Greece and Rome to China to ‘today’s major religions’. The Crusades did demonstrate ‘the rigid boundaries which Christianity, especially Roman Christianity, was creating to mark itself off from all religious Others’ (p. 73), most notably Jews who were ‘massacr[ed] [if they] refused to convert’. But Crusaders showed their devotion to their cause by ‘march[ing] barefoot or fasting before a major battle … Such symbolic links with divine providence heightened the crusaders’ confidence …’ (p. 74).

Is this trust? If so, it is only in-group trust. Would God appreciate these symbols? According to whose theology?

Beyond this argument, Hosking’s treatment of religion is curious. He discusses the ‘Jewish-Christian tradition’ (pp. 56-60). There is more than one Christian tradition, with Catholic teachings and organizations being far more hierarchical than are mainline Protestant ideals. Protestants are divided into believers who hold that the Bible is the literal word of God and cannot be questioned and those who say that it is written by men: this is what separates ‘fundamentalists’ from ‘mainline’ Protestants and the differences are not trivial. Muslims accept the Koran as the literal word of Allah that must be memorized and cannot be challenged. Jews, on the other hand, constantly bicker over what the Torah (the Old Testament) means. There are many interpretations (the oral law of the Talmud and the Mishnah) and even a mystical numerology (Gematria) that reveals what some believe is the true meaning of the holy book. Argumentation is exalted, since the name Israel (which God gave to Jacob) means ‘he who struggles with God’.(7)

Some faiths such as mainline Protestantism and Judaism teach that it is our obligation to treat the stranger like ourselves. This is the foundation of generalized trust.  But for many religions, trust in people like ourselves is the central message. The Crusades were not a historical accident. Nor were other holy wars conducted by members of many faiths. Hosking seems to realize this (p. 148), but his overall vision is that religion promotes trust. In one sense, he is certainly correct: Religious people do more good deeds than those without faith. ‘Mega-churches create social bonds’ (p. 146). But again, these deeds and bonds are largely confined to helping people of their own faith.(8)

Money and Trust

Money, Hosking argues, is another symbol of trust. From ancient Greece to the present, coins, bills, shells, and other trinkets have served as symbols of our commitment to a system of exchange based upon trust. Distrust of money is a sign of ‘the general weakening of trust throughout society’ (pp. 87–9).

The problem, Hosking argues (p. 91), occurs when ‘money completely detaches itself from other symbolic systems [which can be] extremely dangerous’. But money – or more appropriately, the love of money (9) – is the source of our malaise. The trust that is so essential to the daily functioning of our lives has now been weakened by ‘manias and panics’ stemming from problems inherent in our capitalist system. ‘We tend to trust beyond – and sometimes well beyond – the point at which sober reason would suggest we should be more cautious’ (p. 105). To whom – or what – do we give this undeserved trust? To each other, to our governments, to the symbols that the money represent, to the capitalist system? We don’t know. But the evidence is unclear as well. Hosking’s argument (p. 106) that euphoria leads to a spike in trust before an economic bust is not supported by data. Nor is there strong evidence that trust fell more generally. There was no significant fall in trust in other people in the United States in the wake of the 2008 financial crisis. Trust in government fell marginally from its already very low level. The big shift was for trust in finance and banks.(10) To argue that ‘trust’ fell is not helpful – nor is it accurate.

Trust, the State, and the Law

The state and the law are also critical elements in building trust, Hosking argues. Without a strong state and firm laws, trust cannot be sustained. In the Middle Ages, informal groups of people would organize to punish wrong-doers, since there was no strong state and thus little trust. But the Middle Ages were also marked by sharp inequality and rigid classes. On Seligman’s account, this economic stratification made trust impossible. Hosking (p. 28) dismisses this, arguing (incorrectly) that Seligman sets a sharp dividing line between periods of economic domination.

But Hosking argues that the law is the source of trust. Quite the contrary. In a highly trusting society, you don’t need to resort to the strong arm of the law to achieve cooperation or even less crime. You can’t create trust by suing people or putting them in jail. Coercion, Diego Gambetta argues, ‘falls short of being an adequate alternative to trust … It introduces an asymmetry which disposes of mutual trust and promotes instead power and resentment’.(11) It is easy to confuse compliance with voluntary acceptance, to confuse the law abiding people of Singapore (where you can get arrested for chewing gum in public) with those of Sweden (with its high level of trust).                                                                    

Reprise

I really wanted to like this book and to learn much from it. I found Seligman’s historical account of trust to be compelling, in part because it has a coherent argument and in part because it has a compelling thesis. No single account of trust can cover all of history. But Hosking’s argument is too wide-ranging. At the end of the day, the reader does not come away with a clear idea of what trust is (or what different forms of trust might be) nor with how it has changed over time – or how history can inform how we conceptualize trust today.

Hosking’s (pp. 24-25) rationale for not heeding the contemporary literature is that ‘… social scientists have written a good deal about trust, but their work has largely been inconclusive’, because they seek to ‘produce a theory applicable to all societies’, that they focus on ‘cases where an individual has to make a conscious choice about whether to trust a particular individual or institution’, that ‘trust is voluntary’, and that it is ‘a good in itself’ so that trust in the untrustworthy leads to ‘social breakdown.’

These unnamed social scientists seem to be doing what social scientists do. They seek generalizations about why trust is higher in some societies and lower in others.  A good model will be ‘applicable to all societies’, but it will account for variations across countries and cultures. Hosking is not immune from this: his arguments on religion and on the state range from ancient Greece and Rome to the Falun Gong in contemporary China and the worldwide economic crisis of 2008.

Is trust always (or even usually) voluntary? Some forms of trust are. How could trust in institutions be other than a conscious choice? But other types of trust may not be so calculated. Generalized trust, on my account, is a value that guides our behavior – impelling us to treat people as if they were trustworthy rather than based on what happened to you yesterday. You don’t trust the untrustworthy, but a trusting person will dismiss random bad experiences rather than lose faith in others because some of their peers behave badly. Social breakdown is far more likely in low trusting societies. Failed states, according to the index constructed by the Fund for Peace, have barely more than half as many trusting citizens as countries that are more stable (although the direction of causality is difficult to establish).

There is evidence out there. Historians should heed it as much as social scientists do. We need to trust each other a bit more.

Notes

1. Adam Seligman, The Problem of Trust (Princeton, NJ, 1997).Back to (1)
2. Eric M. Uslaner, The Moral Foundations of Trust (New York, NY, 2002).Back to (2)
3. Ibid, chapters six to eight.Back to (3)
4. Ibid, chapter three.Back to (4)
5. William O’Neill, American High, (New York, NY, 1986).Back to (5)
6. http://www.people-press.org/2014/11/13/public-trust-in-government/> [accessed 19 December 2014].Back to (6)
7. http://www.myjewishlearning.com/texts/Bible/Weekly_Torah_Portion/family-vayishlah-2.shtml> [accessed 19 December 2014]. (And for the record, Jacob not only struggled with God, but he was declared the victor <http://www.mishpacha.org/wrestling.shtml[accessed 19 December 2014).Back to (7)
8. Or in attempts to convert outsiders; The Moral Foundations of Trust, chapter seven.Back to (8)
9. 1 Timothy 6:10.Back to (9)
10. http://www.gallup.com/poll/163073/americans-confidence-banks-first-time-years.aspx> [accessed 19 December 2014].Back to (10)
11. Diego Gambetta, ‘Can we trust trust?’, in Trust: Making and Breaking Cooperative Relations, ed. Diego Gambetta (Oxford, 1988), p. 220.Back to (11)