Democracies require trust. The basic reason is straightforward: Democratic institutions are mechanisms for making decisions together. But we have little reason to participate if we believe those institutions are rigged against us or our fellow decisionmakers are incompetent or disingenuous. So a healthy democracy will give citizens confidence in the reliability of both their peers and the institutions that register and implement their decisions.

It’s therefore an urgent problem that political trust is eroding rapidly across the globe. The change has been perhaps most acute among established democracies, but of course they have the farthest to fall. The Organization for Economic Cooperation and Development (OECD) reports that trust in democratic institutions has deteriorated across the developed world, with fewer than four in ten citizens professing moderate-to-high levels of trust in nearly every North American and European country. The United States is a bellwether. From 1999 to the present, the number of Americans who say they trust the government always or most of the time dropped from 39 to 19 percent. One out of eight citizens say they never trust government, a position virtually no one claimed prior to the 2010s. A related measure—what political scientists call “external political efficacy”—taps into perceptions of government’s responsiveness and openness by asking citizens to agree or disagree with statements such as “people like me don’t have a say in government” or “I don’t think public officials care much what people like me think.” According to the American National Election Study, after a half-century decline, just a third of the American population scores at a modestly high level on a standard efficacy index; well over half the population was in that category as late as the mid-1980s. The numbers are particularly alarming in longer comparative perspective. High trust in government and efficacy routinely reached three-quarters of public opinion in the 1960s, and in the history of modern polling the percentages have never been lower than today, even at the height of the Watergate scandal. And the trends in political trust extend beyond government to among citizens. The Pew Research Center finds that today only a third of the American public has a “great” or “good” deal of trust in the political competence of fellow citizens; that’s down from 64 percent in 1999.

Elites in liberal democracies have struggled to find a response to this crisis of political trust, especially in the face of recent waves of populist and nationalist sentiment sweeping Europe and North America. But they can’t simply blame external forces; part of their challenge is internal to the liberal project itself. While we might celebrate liberalism’s emphasis on electoral democracy and individual rights, countless theorists and social scientists argue that liberal theory does not give a satisfying account of the liberal democratic citizen, including the vital question of how citizens come to trust democratic institutions and their citizen-peers. The internal inadequacies of the liberal account have led many social observers to search for a supplemental vision of the citizen among the social networks of civil society. On this vision, liberalism provides a thin framework of process and rights, but the work of building thicker civic dispositions—the necessary “social capital” of democratic citizenship— falls to the houses of worship, sports leagues, fraternal organizations, schools, and other networks that exist outside the state and market. Alexis de Tocqueville, the nineteenth-century French social theorist, was a key progenitor of this vision. Thinkers and practitioners from Robert Putnam to Vaclav Havel are its latter-day interpreters.

Those who embrace this vision, including myriad Christian scholars and opinion leaders, often tell a story of civil society as a hero to liberalism or a refuge from it (or often both; the positions are complementary). The civil-society-as-hero narrative suggests that voluntary associations nurture trust and other dispositions of good citizenship that liberal democracies need yet fail to provide; they are “seedbeds of virtue,” as Mary Ann Glendon and David Blankenhorn titled their edited volume on the subject in the mid-1990s. The civilsociety- as-refuge narrative envisions those same associations as a bulwark, a way for ordinary citizens to use their collective strength to push against the tyrannical impulses of government. Richard John Neuhaus and Peter Berger provide perhaps the most influential case in To Empower People (first published in essay form in 1977, then republished as part of a book in 1996), which argued that civil society plays a crucial “mediating” role between individuals and the liberal state.

These are compelling narratives. But I would add a cautionary chapter. Liberal democracy may very well need civil society, but not necessarily because civil society nurtures political trust. Proponents of the heroic story often overstate the political relevance of their position, partly because they misread the social science on dispositions such as trust and reciprocity. And proponents of the refuge narrative often understate the political relevance of their position, partly because their healthy suspicion of the state can have the unintended double-edge of diminishing trust in state institutions. In any case, political trust is not a mere creature of civil society. We cannot fully address the current crisis in liberal democracies by assuming that political trust will flow inevitably out of the ordinary work of churches and soccer leagues.

Trust(ing) in civil society

Trust is a prediction that another person or institution is a reliable partner. While we can impose that reliability through coercive threat—for example, the sanctions of government— the vast majority of collective decisions operate through “social trust,” our expectation of the reliability of ordinary persons or non-state institutions. The examples of social trust fill our daily lives: expecting that friends will show up for a scheduled dinner date, leaving children in the hands of neighbourhood carpools, taking on a social-service project with fellow members at church, handing off budget reports for review to a subordinate at work, making coffee every morning for a spouse—the list goes on and on.

Part of the reason social trust works is the confirming role of our past experiences. We rarely trust—and should not trust!— when others are not worthy of trust. Trustworthiness builds through repeated interactions with reliable partners. But it is also the case that trust moves from the persons and institutions we know to those we don’t. As we go about our daily lives, we develop a disposition of “generalized” social trust in which we see most strangers as trustworthy most of the time. This kind of trust is largely invisible to us, but we experience it every time we drive down the road or buy something at the corner store. We rarely think, “I can trust the oncoming driver or the local retailer because the government will keep them in line.” Instead we operate with an implicit assumption, built over a lifetime of experience, that our interaction with the other drivers or retailers benefits from our mutual collaboration, even if they are strangers. Generalized social trust is crucial because it reduces the costs of social transactions. Imagine if we needed a legally enforceable contract or a formal monitoring regime for scheduling dinner dates, organizing neighbourhood carpools, working together at church, or making coffee for a loved one. It’s no wonder societies with higher levels of social trust tend to perform better economically and their citizens score higher on measures of psychological satisfaction. They are simply more efficient and less aggravating. One might even say that citizens in high-trust societies enjoy greater freedom.

A citizen’s trusting disposition, then, is a great societal asset. But some of the best minds in contemporary social science see it as an asset in decline, part of a broader erosion of key civic norms and networks over the past few decades. Like many others in the 1980s, my first exposure to this scholarly alarm bell was Habits of the Heart, the bestselling study that describes an American civic culture struggling to overcome the liberal temptation to valorize individual preferences at the expense of the broader community. The sociologist Robert Bellah and his research team borrowed the book’s title from Democracy in America, Tocqueville’s magisterial work, in which he uses the phrase to name the patterns of behaviour and thought—the “habits”— that define the national character. Bellah and crew also drew inspiration from Tocqueville in articulating their primary concern: individualism, a habit that was especially unhealthy for the body politic in its modern therapeutic forms. (One of their interviewees, Sheila, famously described her worldview as “Sheilaism,” with a primary tenet “to love yourself and be gentle with yourself.”)

In the 1990s, Robert Putnam followed this thread, but used the metaphor of “social capital” to reconceptualize the habits of civic culture. He found mounting evidence across the globe for collapse in the norms that hold public life together, including the social trust that individualism tends to diminish. His studies in Italy (Making Democracy Work) and the United States (Bowling Alone) were on the leading edge of renewed scholarly attention to the condition of the liberal democratic citizen.

Much of that academic work has focused on the role of our interactions in groups, variously described as “community,” “networks,” or—as I shorthand it—”civil society.” One of the concerns for social trust is simply the decreasing volume of activity within civil society; we don’t develop trust in isolation, so less interaction in groups is a threat to trust. Putnam’s evocative title in Bowling Alone makes the point: We are engaging each other in civil society less and less, in part because there are fewer opportunities for engagement—fewer sports leagues (we bowl, but we do it “alone”), churches, labour unions, fraternal organizations, volunteer-based social services.

Religion and local community, which Tocqueville identified nearly two centuries ago as key features of American civil society, represent some of the more troubling examples of recent decline. Faith-based associations still occupy a central place in civil society in nearly every Western democracy, with an exceptionally pronounced role in the United States, where studies show that religion motivates as much as half of all activity in civil society. Yet recent patterns of widespread religious disaffiliation—the rise of the so-called nones—suggest an unsettled future for civil society, at best, or an unfilled void, at worst. And while new movements in urban renewal show promise for rebuilding social capital through local associations, they still tend to be limited to certain economically advantaged pockets within a few cities. Moreover, the hollowing out of rural American towns has recently come into sharper view, especially in the wake of the 2016 presidential election. As Katherine Cramer Walsh (The Politics of Resentment), Arlie Russell Hochschild (Strangers in Their Own Land), and J.D. Vance (Hillbilly Elegy) have chronicled, local associations in those communities— churches, small business groups, informal networks of neighbours—have crumbled, and with them the social trust they used to cultivate.

But the story of civil society and trust is not only about volume of associations; it also touches on the richness of organizational participation, which is also under profound threat. Theda Skocpol, in her Diminished Democracy, argues that groups during the twentieth century moved from a model of grassroots leadership to professionalized management of organizations. Groups no longer include their own members in defining mission and vision or even in doing the basic work of the organizations themselves. That is now the job of staff who are paid for their competence. The role of the member has become consumptive, not developmental: We enjoy the selective benefits a group sells by paying a simple fee, not with the shared energy of volunteered labour. The Elks and Kiwanis are old news. The new organizational paragon is the AARP, to which nearly forty million members pay less than twenty dollars a year for everything from glossy magazines to 20 percent off at the Ramada Inn.

Leaders at the AARP might protest that the group does provide opportunities for solidarity. Members are invited to vacation together or join online discussion groups, for example. But set aside the question about whether a cruise or a comment board are really incubators of social trust and reciprocity. The concern is that the activity itself is designed for members only. This points to a broader and growing problem in our contemporary experience of groups: its increasing segmentation and polarization. Our sorting into groups of like mind, driven by everything from housing choices to Internet use, has a peculiar influence on social trust. Polarization does not necessarily diminish trust, but on the contrary amplifies it among those who share group attachments. Putnam calls this the difference between “bonding” and “bridging” social capital. We bond intensely with those within our tribe—our “in-group”—which can exacerbate a visceral distrust of those on the outside.

And it is not enough to say that simple exposure to diversity will help rebuild the bridges. In a finding that holds across Western democracies, Putnam and many others have noted that higher levels of ethnic diversity in a region are actually associated with lower levels of social trust, as groups eye each other suspiciously across their lines of difference. Bridging capital requires serious “cross-cutting networks,” as political scientist Diana Mutz puts it in Hearing the Other Side, in which diverse groups of people are not simply exposed to differences but also make collective decisions together. The rub is that this kind of deliberative decision-making can reduce participation, because citizens who confront diverse ideas are often paralyzed by moral complexity.

Bringing politics back in

The upshot of this scholarship is that civil society is increasingly straining to cultivate the healthy habits of the liberal democratic heart. So a response might seem to follow: We need to re-energize civil society. Ideas for just such a revival have proliferated over the past few decades, with proposals ranging from the relatively straightforward (e.g., increase the earned income tax credit to strengthen disadvantaged families) to the complex and comprehensive (e.g., wholesale revision of urban planning codes to increase diversity in neighbourhoods). It is well beyond the scope of this essay to inventory them, let alone provide an assessment. But it is worth noting that if the goal of a revival is strengthening the qualities of good citizens, including their engagement with and disposition toward the state, our confidence in civil society might be misplaced. Trust in government and fellow citizens qua citizens is a distinctive challenge for liberal democracies that civil society is not fully equipped to address.

In the first season of The Crown, Netflix’s rich biographical drama about Queen Elizabeth II, the British government is confronted with an international crisis in the wake of the Soviets’ test of a hydrogen bomb. But the nation’s key political leaders— Winston Churchill, the legendary prime minister, and foreign minister Anthony Eden—were suffering from incapacitating health crises. To maintain an illusion of strength, and to insulate the young queen from complicated matters of state, Churchill conspires to hide their illnesses from Her Majesty. She inevitably catches wind of their deception and summons Churchill for a scolding:

You were at my coronation, and you therefore heard for yourself as I took the solemn oath to govern the people of my realms according to their respective laws and customs. Now one of those customs is that their elected prime minister should be of reasonably sound body and mind. . . . But it seems that you have not been of sound body and mind these past weeks, and that you chose to withhold that information from me—a decision which feels like a betrayal, not just of the covenant of trust between us and the institutions that we both represent, but of our own personal relationship.

The Queen understands the personal affront partly in interpersonal terms, a violation of social trust born out of her regular interactions with Churchill. She trusted him because she had taken stock of his dependability and character. Much of the social-capital literature I have discussed taps this way of thinking. Trust builds “exogenously” to political institutions; it does not require the intervention of the external authority of the state.

But the Queen also invokes an institutional explanation for the trust she placed in Churchill, a confidence rooted in a “covenant” that includes rights, rules, offices, political traditions, and sanctions. She trusted him against a backdrop of British constitutionalism, with its emphasis on the rule of law and the respective roles of head of state and head of government. Social scientists often say that social trust in this sense is “endogenous”—or internal to—political institutions. But note that the Queen does not merely identify two distinct sources of trust, the social and political; she envisions them as complementary. After all, her personal assessments of the man were largely observations of his role in a political office.

Of course, this is a rarefied relationship defined by politics. But consider more mundane examples. I do generally trust drivers on the road without conscious thought to rules and sanctions (social trust), but that confidence has been partly habituated through the government’s placement of stop signs, road dividers, and occasional speed traps. I do generally trust the proprietors of the corner store without worrying that they might cheat me (again, social trust), but that confidence has been partly habituated through the government’s mechanisms for consumer protection. In reality my expectations about ordinary people and my expectations about political institutions are not mutually exclusive sources of trust; they are distinct yet mutually reinforcing.

What does this have to do with reviving trust in liberal democracy? Even though social and political trust interrelate, the effect of one on the other is not as clear as civil-society-as-hero narratives can sometimes suggest. Civil society can breed social trust, to be sure, and sometimes it can even foster trust that bridges across lines of difference, but social trust does not necessarily affect levels of trust in government or in fellow citizens (in contrast to co-parishioners, family members, co-workers, friends). I might trust my neighbours or co-religionists yet see government as hopelessly corrupt or non-responsive. I might also trust my neighbours and co-religionists in those roles—as dinner mates, carpool drivers, investment advisors, maybe even spiritual mentors—but find them incompetent or even morally dubious as citizens. These are hypotheticals, but in fact the empirical evidence that social trust breeds trust in government is largely non-existent. Some studies even suggest the influence might flow more clearly in the other direction: Political trust—confidence in the reliability, openness, responsiveness, and fairness of government— often acts as a precondition for social trust, not the other way around.

To nurture political trust, we need to take seriously that it relates citizens to the state. The key problem, as I see it, is that civilsociety narratives often neglect that distinctive citizen-state relationship. This problem is nowhere more pronounced than contemporary trends in civic education. It is common in secondary and higher education, for example, to introduce students to citizenship by requiring that they serve with social agencies or other groups in civil society. The hope is that volunteerism of this kind will generate greater civic-mindedness later in life—and it often does. But the implication is that “civic” is best understood in narrow terms that have little to do with the state. On the one hand, young people actively volunteer in ways that do some good for society and themselves; in that sense, as political scientist Russell Dalton argues optimistically in The Good Citizen, they are “engaged” citizens. On the other hand, the data on voting, party affiliation, interest group membership, political protest, participation on local boards and commissions, political knowledge—all the nittygritty of conventional democratic politics— suggest that these same young people see government with impatience or indifference. They simply do not trust the state’s potential as a site for collective action in pursuit of public justice.

I am not proposing that our relation to the state encompasses the whole of democratic citizenship. Nor am I making a veiled critique of sphere sovereignty, subsidiarity, or other theological constructs that emphasize the important public role of civil society. We do act as citizens when we volunteer within civil society in service to the common good—and that engagement requires an interpersonal trust that civil society helps to build. But we will not generate trust in that other “sphere” for citizenship—the state—if we treat our relationship to the state as just another species of those ordinary relationships we develop in civil society. The citizen-state relationship is clearly different, not least because it entails both an institution that uses violence and obligations that we cannot avoid (we can quit voluntary associations; we can’t “quit” the state).

The crisis of political trust is not rooted in a problem of “too much politics,” as we often hear in response to government’s many dysfunctions. Rather, the crisis is rooted in too little attention to political capacities that citizens use to engage each other as citizens. Those capacities are learned through civic experience and civic education that our ordinary interactions in civil society rarely provide. A revival of political trust, then, requires a distinctively political account of the citizen.