WonkBlog reports:

Three quarters of whites don’t have any non-white friends

“All my black friends have a bunch of white friends. And all my white friends have one black friend.”

That’s the memorable punchline of a Chris Rock bit from 2009 on interracial friendships. And according to some recent number-crunching by Robert Jones of the Public Religion Research Institute, there’s a good deal of truth to that statement.

Let’s consider the average American white American and the average American black American, and let’s say, for simplicity’s sake, that each of them have 100 friends. If you were to break down their respective friend networks by race, they would look something like this.

In a 100-friend scenario, the average white person has 91 white friends; one each of black, Latino, Asian, mixed race, and other races; and three friends of unknown race. The average black person, on the other hand, has 83 black friends, eight white friends, two Latino friends, zero Asian friends, three mixed race friends, one other race friend, and four friends of unknown race.

Going back to Chris Rock’s point, the average black person’s friend network is 8 percent white, but the average white person’s network is only 1 percent black. To put it another way: Blacks have ten times as many black friends as white friends. But white Americans have an astonishing 91 times as many white friends as black friends….

But wait: Just how were the friend counts obtained? At the bottom of the post, we see that things might be a bit complicated: “As part of their American Values Survey, PRRI researchers asked respondents to name up to seven people with whom they regularly discussed important matters. They then asked a battery of demographic questions about these people — their relationships to their respondents, as well as their gender, religion, and germane for these purposes, their race. They used these numbers to derive average racial breakdowns of the friend networks of the average black, white and Hispanic survey respondent.” So the underlying survey only asked respondents to name up to seven people, presumably the closest of friends, and not necessarily reflective of all their friends.

But beyond this, if one looks at the survey results, one finds that

  1. the median number of people about whom respondents were talking was just three (average 3.4), and
  2. many of those people were family members — spouses or partners, parents, children, and siblings.

The author of the Wonkblog post kindly got me in touch with the researcher in charge of the survey, and I got some more data; and it turned out that the average respondent only named 1 or 2 people outside his or her immediate family. (The average was 1.5.)

The study thus doesn’t mean, I think, that “three quarters of whites don’t have any non-white friends,” under any normal unmodified definition of “friends.” Rather, it is (more or less) that “three quarters of whites don’t have any non-whites among their three or four closest confidantes, including family.” Or, if you prefer, one can rephrase it as, “one quarter of whites has a non-white as a close family member, or within their one or two closest confidantes.” (I say “more or less” because this doesn’t map precisely on the study results, since some people reported no confidantes at all, and some reported seven — three is just the median.)

And it’s also hard to see how one extrapolate from that to “friends” generally, including to a hypothetical 100 friends. Consider your own friend network: Are the three people you know best reflective of the 100 people you know best, whether by race, religion, social class, profession, and so on? My guess is that we tend to see more homogeneity along many dimensions in our closest circles than we do in our broader circles.

But in any event, I just don’t see how one can say much about the 100-friend circle based on what we know of the one-or-two-friend circle (or the three-friends-and-family circle). And “white Americans have an astonishing 91 times as many white friends as black friends” seems to me an especially difficult estimate to justify.

Now it may well be that these close family-and-friends circles are especially important, more so than the total circle of friends, in influencing how our views are shaped and whom we most empathize with. And it may be that whites (and blacks, as the study suggests) don’t have a lot of friends (even going outside the family and one or two confidantes) of other races. It’s just that the study doesn’t really demonstrate this second point, and “three quarters of whites don’t have any non-white friends” strikes me as an unsound way of conveying the first point.

Or am I missing something? If I misunderstood the study, or the report of the study, please let me know.