Robert Putnam warned nearly two decades ago about what he called the "technological transformation of leisure." The influential academic, who went on to pen the book Bowling Alone, put it this way back in 1995:
There is reason to believe that deep-seated technological trends are radically 'privatizing' or 'individualizing' our use of leisure time and thus disrupting many opportunities for social-capital formation.
In other words, the more people sit at home, alone, in front of a glowing screen, the less time we have to form the kind of social bonds with neighbors, friends and organizations that help strengthen communities. Putnam was thinking 18 years ago about a deep-seated technological trend that sounds almost outdated today: "The most obvious and probably the most powerful instrument of this revolution," he wrote, "is television."
The 2013 instrument is, of course, the computer (or the laptop, or tablet, or even the smart phone). And, in fact, following Putnam's lead, there has been much debate about what the Internet may be doing to social capital now that our private, glowing screens can simultaneously be networked to each other. On the one hand, opportunities for you to technologically amuse yourself at home alone have proliferated since 1995. On the other, research from the Pew Internet & American Life Project suggests that people who are active online actually have larger social networks and are more likely than non-Internet users to participate in social groups (online and in real life).
So is the Internet getting in the way of our social lives, or enhancing them in new ways? And given Putnam's thesis that social capital matters for a vibrant civil society, should we be worried about any decline in, well, actual face-to-face human interaction?
Enter a new National Bureau of Economic Research working paper by Scott Wallsten with the amusing title, "What Are We Not Doing When We're Online." Wallsten, an economist at the Technology Policy Institute, looked at 2003-2011 data from the American Time Use Survey in an effort to quantify which activities we're short-changing as we devote more leisure time to our computers.
In some cases, we've just changed how we do the same things we were already doing (we read the newspaper online instead of spreading it out in print). In other cases, as with Facebook and Twitter, there is no obvious pre-Internet analog. Time we kill on Facebook is time we probably previously spent doing wholly unrelated activities, like sleeping more, or eating out with friends.
The American Time Use Survey suggests that the total amount of time we spend on leisure has remained relatively constant between 2003-2011, at about five hours a day. The subset of that devoted to Internet leisure has been increasing, and is now about 13 minutes a day by the ATUS definition (Wallsten: "This average is deceptively low in part not just because it does not include time spent doing email, watching videos, and gaming, but also because it is calculated across the entire population").
That said, Wallsten has calculated that each minute of leisure we spend online correlates with 0.29 minutes we don't spend on other types of leisure, like watching TV and going out. One added minute online correlates with 0.27 fewer minutes working, and 0.12 fewer minutes sleeping. Via Wallsten:
My results also suggest that other offline leisure activities that involve interacting with other people are crowded out by online leisure: attending parties and attending cultural events and going to museums are all negatively correlated with online leisure. In short, these results based on ATUS data suggest that a cost of online activity is less time spent with other people.
That is the kicker that brings us back to Putnam. Now, the Internet is different from the original culprit, the TV, in that it contains as many social corners as solitary ones. It's possible that the social connections you make on Twitter in some way compensate for the ones you might have made at that party. Or perhaps those online social connections are of an entirely different nature, with social costs or benefits that we haven't figured out yet.
Wallsten makes no value judgment on this, but he argues that the crowd-out effects that he's observed are large enough that we should consider them among the true economic effects of the Internet:
While I control for a large number of relevant factors in the analyses, the relationships between online and offline time are correlations, meaning we cannot say definitively that an incremental minute translates into a tenth-of-a-minute less sleep. Perhaps, instead, when people suffer from bouts of insomnia they take to the Internet, either to look for insomnia cures or other ways of passing a sleepless night. Nevertheless, the analysis shows that online activities, even when free from monetary transactions, are not free from opportunity cost.