ON CHRISTMAS DAY 2004, Rick Warren, the pastor of Saddleback church in southern California, got a telephone call from Thailand. The caller, a pastor whom Mr Warren had trained, said a tsunami had just been reported and many people would need help. Before the tsunami had even struck Sri Lanka, Mr Warren was calling a vast network of churches in South-East Asia who got parishioners to safety, and e-mailing an even vaster network of clubs in southern California. Thousands of volunteers went into action overnight. Within a day, food and medicine worth millions of dollars were winging their way from a single church to the disaster-hit regions. This was American civic volunteerism in action, updated for the 21st century.

Voluntary associations have been the secret ingredient of American social dynamism since the country's foundation. “Americans of all ages”, wrote Alexis de Tocqueville, “are forever forming associations.” By connecting people to their neighbours and to the wider world, argued de Tocqueville, civic associations made Americans better informed, safer, richer and better able to govern themselves and create a just and stable society. They developed to provide something for everyone: fraternal bodies such as the Kiwanis and Elks for men, cross-class federations such as the League of Women Voters for women, scouts and girl guides for boys and girls. There were farmers' groups, industrial unions and associations of Elvis impersonators. If the great sorting-out is pulling social bonds apart, then Americans' love of clubs seems the most likely glue to put them back together again.

But will it? In 2000, Robert Putnam, a professor at Harvard, wrote a celebrated book, “Bowling Alone”, which claimed that America's civil society was in crisis. Mr Putnam pointed out that in the two decades to the mid-1990s, the number of Americans who said they had attended a public meeting on local or school affairs had fallen by a third. The proportion who said they attended club meetings fell from nearly two-thirds of the population in the mid-1970s to a little over one-third in the late 1990s.

It was not solely the decline in volunteerism that was the problem. The composition of civic associations was also changing. As many of Mr Putnam's critics pointed out, the absolute number of voluntary associations rocketed during that period, from around 8,000 in the 1950s to just over 20,000 in 2000. Optimists said this showed that, far from falling apart, civil society in America was actually flourishing. Americans were finding new ways of linking up and influencing public life.

However, it turned out that the new associations belonged to a different species. Whereas the old clubs had many functions, social, recreational and professional, most of the new organisations—bodies such as the Sierra Club or NARAL, a reproductive-rights group—were professionally led advocacy groups. Whereas the old clubs had networks of local chapters that met frequently, the new bodies often had no such networks, or their chapters met only rarely. And whereas the old clubs depended on annual membership dues, the advocacy groups got their money from foundation grants or direct-mail appeals.

The result, argues Theda Skocpol of Harvard University, was that mass participatory civic life in America declined, despite the proliferation of new bodies. The old associations reinforced ideals of good citizenship. Local and national leaders had to take some note of their members' views. And surprisingly large numbers learned the workings of democracy at first hand. National federations could have thousands of local chapters, each with a number of officers elected each year. Mrs Skocpol calculates that in 1955, the 20 largest voluntary associations alone had 5% of the adult population serving in some capacity or other. All these local officers had to run meetings, handle membership dues, keep records and so on.

America needs himAP

By contrast, people typically become members of one of the new bodies by replying to a direct-mail shot. There are no local officials. National leaders spend their time fund-raising or lobbying politicians in Washington to change the law. These are not primary schools of democracy.

At first sight, the shift from one kind of organisation to another seems inevitable: it is a product of mobility. “People are like plants,” says Mr Putnam. “When you re-pot us, it takes a while to put down new roots.” Clubs tend to be products of stability. But Americans were moving around a lot, so they started joining long-distance organisations instead of face-to-face clubs.

Yet that cannot be the whole story. There were more face-to-face clubs in the 1950s, when geographical mobility was even higher than it is now. Part of the explanation is that civic volunteerism is also being eroded by things that have little to do with mobility per se: television and computers; hours spent stuck in traffic jams; even the decline of family meals because both parents are at work.

In need of friends

But the bigger part of the explanation has to do with the paradoxical impact of mobility itself on people's clubbiness. Its direct effect is to disrupt community ties. You see this, for example, among Hispanic immigrants, who have lower rates of joining voluntary associations than do native-born Americans. But the indirect effect of mobility is to stimulate the appetite for social bonds, so when they arrive, immigrants strive to rebuild networks—for example, by organising huge celebrations for Cinco de Mayo (Mexico's national day).

Hence the decline of community ties may well be connected to continued geographic mobility. But that does not mean, as pessimists fear, that mobility will continue to eat away at them until America becomes totally atomised. Because mobility stimulates demand for clubs, you would expect to find new forms of civil volunteerism starting to crop up. And you do: after years of decline, civil society is staging a comeback.

This is showing up in three areas. First, after the terrorist attacks of September 11th 2001, civic and public engagement spiked. Church-going and other sorts of voluntary activity increased. Trust in government rose sharply, to a level not seen since the 1960s. Most of these things dropped again soon afterwards, but not all of them.

The 2004 election saw the biggest turnout for almost 40 years. George Bush's vote was 23% higher than it had been in 2000, and John Kerry's was 16% higher than Al Gore's had been, though the smaller parties did less well. The campaign also produced volunteers in droves. There were specific explanations for this: the election was close, and America was at war. Still, September 11th may also have re-engaged Americans with public life.

Moreover, there are now four years of indicators showing that the terrorist attacks had an enduring impact on one specific group: those who were in their late teens or early 20s at the time. A project at the University of Maryland under Bill Galston is looking at the attitudes of this dotcom generation. In the year after September 11th, volunteering by 15-25-year-olds leapt, with 40% saying they had volunteered over the previous 12 months, compared with a third for all age groups. That surge has endured. The University of California at Los Angeles is tracking civic engagement at universities, and has found that college freshmen are much more likely to keep up to date with and discuss politics than they were in 2000.

After the second world war, the shared experience of combat created a “greatest generation” that consistently voted more, joined more clubs and volunteered more hours than either their parents or their children. It is too early to tell whether the terrorist attacks will have anything like that impact. But the evidence is growing that they have created a “9/11 generation”.

Second, the internet is at last beginning to have an effect on voluntary associations. When Mr Putnam published “Bowling Alone”, one of the most common objections to his thesis was that he had underestimated the role of technology (which was then caught up in the dotcom bubble). Yes, the techno-optimists said, people may not be joining traditional clubs, but they are joining “virtual communities” online. Five years later, they are being proved right.

The electronic welcome mat

At first, it seemed as though Mr Putnam's scepticism was justified. The anonymity of virtual communities appeared to be fatal to the creation of trust and of real civic bonds. When half the members of a teenage girls' online chatroom turned out to be middle-aged men, it was hardly a Tocquevillean moment. But over the past few years, online groups have started to use “convening technology” to create face-to-face social bonds. For example, if you go to meetup.com, you can type in where you live and what your interests are—say Young Republicans, Chihuahuas or Brazilian reggae—and the site will tell you where and when Young Republicans, Chihuahua fanciers or Brazilian reggae enthusiasts are meeting up within 15 miles of your home over the next two weeks.

One of the founders of Meetup, in New York in the aftermath of the September 11th attacks, was Scott Heiferman, who had been reading “Bowling Alone”. “In those first few days after 9/11”, he says, “New York felt more like a city of neighbours than of strangers. The question I started with was: how do you start an association today? Do you need a building in Washington? No, you go online.”

Since 2002, Meetup has been the forum for over 100,000 clubs with 2m members. This spring there were 2,400 Meetup meetings a week, about 50% more than a year earlier. It is true that these are often meetings of the like-minded, of people with particular tastes in common, rather than universal gatherings of the old national federations that tied disparate groups together. It is also too early to say whether a single website will reverse decades of civic decline: meetup.com is only three years old and has just started charging, with unpredictable consequences.

Still, the growth is impressive. And Meetup is not the only example. One of the most interesting sideshows of the 2004 election was Howard Dean's successful use of the internet to mobilise supporters and raise money. Politically, his campaign was a disaster, but party managers everywhere took notice.

Moreover, these internet-generated meetings are the traditional face-to-face kind that Mr Heiferman calls “the old Tocquevillean stuff”. To cite one small example, the New York Chihuahua club organised a protest by city dog owners against attempts to close dog parks. “The 300 Chihuahua groups around the world think of themselves as a chapter-based organisation,” says Mr Heiferman. In the past, when you moved house, you could lose touch with your local chapter. Now you go online and plug into a network.

The third force helping to revive old-fashioned civic life is religion, and one sort of religious organisation in particular, the megachurch. In many ways, religion itself is an exception to the rule that chapter-based voluntary associations are in decline. Churches are the archetypical civic clubs. Congregations meet regularly face to face. They give money every week. Parish committees are the equivalents of local chapters. Churches have long played a hidden (and sometimes not so hidden) part in political life, from the black churches' role in the civil-rights era to Evangelical Protestant churches in Mr Bush's re-election.

Reliable figures on recent church attendance are hard to come by, but both Gallup polls and church records suggest it was flat between about 1980 and 1998. Some studies, and a mass of anecdotal evidence, suggest that in the past few years attendance may well have been rising. It is pretty safe to say that church attendance has not fallen—and that alone makes churches stand out from the general civic decline.

Getty Images

But over the past few years, one particular part of the American religious line-up has done far more than merely avoid decline. This is the megachurch, usually defined as a church with an average congregation of more than 2,000 each weekend. In 1960, there were fewer than ten of these. Now the number has risen to at least 1,200, according to the Hartford Institute for Religion Research. Twenty of these churches have weekly congregations of 10,000; three have over 20,000. They are Willow Creek outside Chicago, Lakewood in Houston and California's Saddleback. Nearly all their growth has taken place in the past 20 years. Saddleback, which has 82,000 people on its church rolls, held its first service 25 years ago.

Mr Putnam points to two aspects of the success of megachurches that matter to civic organisation in America. Neither has anything to do with doctrine. The first is that founders of megachurches made them attractive to people who were mildly religious but not committed in advance. Mr Warren of Saddleback and Bill Hybels of Willow Creek both went round asking neighbours who did not go to church what they disliked about it and what would persuade them to go.

They redesigned their churches on the basis of this consumer research. Out went 200-year-old hymns, pulpits and even church-like buildings. In came information booths, food courts, churches that look like schools, reggae services and sermons with Powerpoint presentations. As one Californian pastor told the Hartford Institute, “We're trying to create an environment here so the unchurched person can come in and say, ‘This is a church like I have never known.’ ” If megachurches can win millions of new supporters by lowering barriers to entry, there is no reason why secular civic organisations cannot do the same.

The second and more important feature of the megachurches is that they have created huge numbers of small groups. Saddleback has 2,600 small groups, co-ordinated by 9,200 ministers, for its congregation to join—for computer nerds, cyclists, knitters, you name it. This is a close approximation to the thousands of chapters of the old-fashioned voluntary associations, and makes it possible to engage in social activism on a grand scale. Last year, Saddleback decided to feed all 42,000 homeless people in Orange County for 40 days, and did it.

According to Peter Drucker, a veteran management writer, “The most significant phenomenon in the first half of the 20th century was the company. The most significant phenomenon in the second half of the 20th century was the large pastoral church.” Just as small churches civilised the Wild West because they were the first real community organisations in frontier towns, so megachurches—which are most prevalent in new suburbs and in the sunbelt—are providing community bonds on the new frontier of America's middle-class migrants.