A century and a half ago, the French journalist Alexis de Tocqueville travelled throughout the U.S. marvelling at the generosity of the American people. "When an American asks for the cooperation of his fellow citizens," Tocqueville wrote, "it is seldom refused; and I have often seen it afforded spontaneously, and with great goodwill" . He attributed this generosity to Americans' tendency to look beyond their own immediate concerns. "The principle of self-interest rightly understood produces no great acts of self-sacrifice, but it suggests daily small acts of self-denial ... If the principle of interest rightly understood were to sway the whole moral world, extraordinary virtues would doubtless be more rare" .
This idea of self-interest (rightly understood) we now call "trust" in other people. According to Tocqueville and many others, people develop trust in each other when they join for common purposes in civic associations. Indeed, Robert Putnam, a professor at Harvard University, refers to a "virtuous circle" of trust, including group membership and informal social ties, that has become known as "social capital" . Social capital helps make society and its governments run more smoothly.
Beginning in the late 1960s and early 1970s, Americans began to withdraw from participation in all sorts of civic groupsfrom such traditional service organizations as the Rotary Clubs, Kiwanis, and League of Women Voters, to bowling leagues and card-playing clubs. Americans socialized less with friends and neighbors and voted less often in elections. The inevitable result was that Americans became less trusting of one another . In 1960, 58% of Americans believed "most people can be trusted" (as opposed to saying "you can't be too careful in dealing with people"). By the 1990s, barely more than 33% of Americans trusted one another, according to national surveys, including the General Social Survey and the American National Election Study. Americans had lost their sense of community. They didn't mix with one another as much as they once did and didn't trust one another as much as they once did. They became more balkanized, and their public life became more contentious, while their national institutions (especially the U.S. Congress) found it increasingly difficult to compromise on even the most basic public policy questions.
The villain in the decline of social capital, Putnam argues, is technologyinitially television, now the Internet . Watching a lot of television keeps Americans inside their homes and away from the civic organizations and social connections that generate trust. Heavy television watching also leads people to imagine the real world is as mean and violent as the programs they seemaking them less likely to trust strangers. Television produces misanthropes who view the world as a dark and threatening place and whose "friends" are fictional characters they'll never be asked to help out.
Today, there are even more mistrusters, and civic engagement has dropped even further. The new culprit seems to be the Internet. Even more than television, the Internet may be a lonely place. We hear of people spending hours in front of computer screens, becoming addicted to the Net, ignoring their families and dissociating themselves from friends. Television programs may make you think the world is mean, but the Internet demonstrates just how nasty people can really be. When people enter an Internet chat room, they can hide their identity, "flame" other people, and "troll" first-time visitors to a Web site. The Net can be a dangerous place, where "charities" solicit funds for nonexistent causes, scoundrels feign love for lonely hearts, and unscrupulous hackers uncover and collect and resell credit card numbers. Heavy Net users become more depressed, lead more stressful lives, and have fewer friendseven though they may start out as well-off psychologically as everyone else .
The Net is neither a dark and threatening place nor a grand intellectual and social commune.
Still, this is just one face of the Internet. Others view the Net as a great opportunity to rebuild our lost sense of community and trust. People come together on the Net through email lists, affinity groups, support groups, and chat rooms. The Internet connects people from all over the worldand may be, as Michael Hauben and Ronda Hauben described it, "a grand intellectual and social commune in the spirit of the collective nature present at the origins of human society" [1, 5]. The Net is the great leveler of class and race barriers in U.S. society [11, 12]. As computer literacy and Internet access increase, Americans should reconnect with one another, thus forming the foundation for a new era of trust.
I offer a third perspective. The Internet neither destroys nor creates social capital. There are altruists, as well as scoundrels, on the Net, just as there are in everyday life. Indeed, the Internet, like television, mirrors everyday life. What people do online is pretty much what they do offline: shop, get sports news and weather, plan vacations, and, most of all, contact people they already know through email. The Net is not a threat. Neither is it Nirvana.
The major reason why the Net is not the "new new thing" of trust and civic engagement is that much of the current discussion of the virtuous circle of trust, civic engagement, and socializing is misplaced. Trust in other people is trust in strangers, or people who are different from ourselves. Trust is essential for a civil and cooperative society but does not depend on our life experienceswhether we are visiting friends and relatives, joining civic organizations, watching television, or surfing the Net. Instead, trust reflects an optimistic world view and belief that others share our fundamental values. We learn trust from our parents. We are not likely to become more trusting in people who are different from ourselves by interacting in clubs or in coffee klatches with people like ourselves .
Trust is far from being irrelevant to the Internet, however. Going online does not make people either more or less trusting, though trust shapes how people interact with one another. People who trust are less likely to fear getting involved with strangers. In everyday life, trusters are less likely to lock their doors at night or use guns to protect themselves. They are more likely to volunteer and give to charity. On the Net, trusting people are likely to view others as nice folks who won't exploit them. They are likely less worried about violations of their privacy and more likely to buy goods online. There is little reason to believe that people who trust others are more or less likely to use the Internet than others.
Perhaps people who use the Internet a lot are hermits, but more likely they are sociable. From email (the most widely used online application), to chat rooms, to support groups, going online involves communicating with other people. But the Internet is not likely to create the kinds of communities that could generate trust. Trust develops between people of divergent backgrounds, whereas the Net excels in bringing together people who already have something in commonfamily ties, friendship, working in the same office, political views , or needing the same kind of medical information or psychological support. One of the most heralded forms of online communitymedical support groupsbrings together people who may know nothing about one another, except that they share the same malady. And since many such conditions are temporary, such as sports injuries, these online communities may well be populated primarily by transients . Nevertheless, simply going online for information and support is hardly the hallmark of a troglodyte.
Trust, Sociability, and Internet Use
What connects trust, sociability, and Internet use? I analyzed data from a 1998 survey of technology use conducted by the Pew Center for The People and The Press (Washington, D.C.). The survey, still the most current publicly available one on Internet use, asked 2,000 Americans a variety of questions about going online, as well as about their social networks and trust in others and in government. (Alas, there were no questions on group membership, a key aspect of social capital.) I estimated almost 20 probit models using a statistical technique called "ordered probit analysis."1 These models allowed me to determine which factors best predict different forms of Internet use. Each model contains many factors that might lead to more Internet use,2 though I focused on trust and measures of sociability (the scope of people's social support networks, how often they visit family members, and how frequently they call friends). Unlike another 1998 survey (by Robert Kraut et al. of Carnegie Mellon University in Pittsburgh) of people who were given computers so they would go online , the Pew Center survey was a representative national sample of the population with a much larger number of respondents2,000 compared to 169. Examining only those people who go online may lead to erroneous conclusions, especially since most people don't spend a lot of time there.
Variables of Internet use fall into four categories: general use; how often users go online; making social connections online; and worries about privacy and security (see Table 1). These results are telling. The Internet is neither the tool of the devil nor the new Jerusalem heralding the renaissance of a national sense of community and trust in one another. For most types of general usegetting and receiving email, getting information on health, business, sports, and stocks, expressing personal views online, and buying goods onlinetrust either doesn't matter at all or doesn't matter much.
Surprisingly, email users are more likely to trust others, and people who buy goods online are ever so slightly more likely to trust others. Beyond that, general use of the Internet is connected neither to trust nor to sociability. All sorts of people go online to seek informationthe trusting and the misanthrope, the sociable and the recluse.
There is little reason to expect that simply going online taps or drains sociability (or trust). Kraut et al. argued that people spending a lot of time online are the misanthropes . But the Pew Center survey offers little support for this view. People who use their computers a lot, who might report they spent a lot of time online yesterday (in real time and how often they connected), and who say they spend too much time on the Net, are no less trusting than people who don't go online at all. Neither the Internet nor television remakes people's personalities . The picture of heavy Net surfers as loners is also wrong; the heaviest users have wider social circles and support networks.
The Net is another outlet for people already connected to other people, as others have also found . Chat roomsa Net innovationoffers some hope that people of different backgrounds might get together and learn to trust one another. But there, of all places, we see evidence of misanthropy. People who visit chat rooms or make new friends online are no more or less sociable than anyone else. They don't have bigger or smaller support networks and are no more likely to visit relatives or call friends. Yet they are less trusting than others.
Perhaps people who make friends online, often anonymously, feel uncomfortable meeting "real" strangers. And many, maybe most, chat rooms are marked by a dominant world view or ideology; dissidents often find out rather rudely they are not welcome [2, 12]. People who frequent chat rooms seem to trust only people like themselves and fear those with different views.
People who mistrust others fear the Net as much as they accept all sorts of other conspiracy theories we might see on the X Files. They worry about their privacy generally and about the security of their medical records and the risk of downloading viruses in particular. Trusters view the Internet as more benign. Trusting people believe they can control the world and have faith that science will solve their problems and that the Net is another tool giving them leverage over their world [10, 12].
The Internet, then, is not a reservoir of social capital. As in everyday life, there are places where trust matters, and there are even more places where it doesn't. Trust matters most when people fear the unknown and worry this new technology could come back to haunt them. There is little evidence that the Internet will create new communities to make up for the decline in civic engagement over the past four decades in the U.S. Still, there is even less evidence that the Internet is pushing people away from traditional social ties or making them less trusting.
Internet use neither consumes nor produces trust. People are no more trusting because of the number of listservs they belong to, how often they go online, whether they are willing to share their views on political or social issues online, or whether they enter chat rooms. In fact, the more people use email, the less trusting they seem to become. Surfing the Net does not turn a misanthrope into a truster.
The 1998 Pew Center survey does not allow a good test of the effects of Internet use on trust, because it did not include most of the factors I have elsewhere found to be the most important reasons people trust others, including a general sense of optimism and control. The 1996 American National Election Survey does permit a rough test of the effect of Internet use on trust. The test is not ideal, because the survey asked only whether people had Internet access, not whether they went online or what sites they visited. Still, it is the best test available, showing just what I would expect: Internet access leads to neither more nor less trust, once we account for other factors, including general optimism, age, and education, among other variables.
Circling the Wagons
This conclusion completes the circle. The Net is neither vicious nor virtuous, neither a dark and threatening place nor a grand intellectual and social commune. Yes, the Internet is filled with pornography, but it didn't invent it, and nobody is forced to visit these sites (or others that sell Viagra to dogs, promote worthless stocks and submerged real estate, or encourage children to gamble). And, yes, there are plenty of opportunities on the Web to give to charities, find volunteering opportunities, and join support groups.
But even this combination of good and bad intent isn't the whole Internet either. The Web is very much like the physical world, making things better in some ways and worse in others. It is not transforming. If you want to make a revolution, go offline. The message of the findings in the Net-use surveys is that the Net is neither a threat to our society nor to its moral fiber. Regulating it won't solve our social problems or save our children from evil influences. Children develop trust in others by learning from and emulating their parents, not from what they (don't) see on television or online. How much people trust one another as children largely determines how much they trust people as adults. The world may seem a more dangerous place when viewed through the prism of television and the Web. And the Internet makes truly mean-spirited sites more readily available than they would be in the everyday world (or even on television).
The Net-based ideological free-for-all does not mean that the Net (or any other medium) poses a threat to most families. By itself, it represents neither a threat to civil society and sociability nor its panacea.
2I use other variables to ensure the relationships between Internet use, on the one hand, and trust and sociability, on the other, are not spurious. The other variables are: age, gender, being a student, family income, home ownership, being self-employed, being single, having no religion, time spent watching television each day, time spent reading a newspaper, trust in the federal government, and belief that the Internet helps people keep in touch with one another.
The General Research Board of the University of Maryland, College Park, facilitated this research through a Distinguished University Research Fellowship.
©2000 ACM 0002-0782/00/1200 $5.00
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. To copy otherwise, to republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee.
The Digital Library is published by the Association for Computing Machinery. Copyright © 2000 ACM, Inc.