"Americans have a third fewer close friends and confidants than just two decades ago."

USAToday reports:

In 1985, the average American had three people in whom to confide matters that were important to them, says a study in today's American Sociological Review. In 2004, that number dropped to two, and one in four had no close confidants at all....

The study finds fewer contacts are from clubs and neighbors; people are relying more on family, a phenomenon documented in the 2000 book Bowling Alone by Robert Putnam, a Harvard public policy professor.

The percentage of people who confide only in family increased from 57% to 80%, and the number who depend totally on a spouse is up from 5% to 9%, the study found.

I'm surprised the number who confide only in family is so high, but then why is the number who confide only in their spouse so low? I guess that 80% who confide only in family confide in more than one person, often including a spouse. You can't tell from those statistics what percent of individuals don't confide in their spouses or who confide in only one family member other than a spouse. What percent of Americans confide only in their mothers? More than 9%?

Why people have fewer close friends is unclear, Putnam says. "This is a mystery like Murder on the Orient Express, in which there are multiple culprits."

The chief suspects: More people live in the suburbs and spend more time at work, Putnam says, leaving less time to socialize or join groups.

Also, people have more entertainment tools such as TV, iPods and computers, so they can stay home and tune out.

Yes, what about those blogger characters who confide in the whole world? That's not me. But I'm just saying...

A question I have about the statistics -- and I've only read the news report, not the underlying scholarly article -- is whether the idea of what it means to confide in or rely on someone has changed over time. "The study is based on surveys of 1,531 people in 1985 and 1,467 in 2004," and I'm sure that basically the same questions were asked. But people may think about the questions differently now than they did then. And what it means to be close to other people may have changed. For example, you might be more likely to have someone to talk to about your sex life but less likely to have someone to take you to the doctor. Maybe we've come to accept more professional help for certain kind of problems that we used to rely on friends to help us with.

1985, the year of the original survey, was also the year that "Dionne and Friends" had a big hit with the song "That's What Friends Are For." The recording was made to benefit the American Foundation for AIDS Research, and that expressed the notion that "What Friends Are For" is to help you when you're sick and you need comforting and -- let's be frank -- money. But what friends are for is open to question and redefinition. It would be a different sort of study that examined what it means to have friends. You might feel that you have several friends that you can talk to about your deep feelings and aspirations but still think there is no way you'd ever ask them for money or to do your errands when you're in the hospital. That's what family are for.