Nearly every week, the U.S. Census Bureau releases a new set of figures on American families and the living arrangements they have been creating in the past decade. And each time, as the media liaison for a national association of family researchers, I'm bombarded with telephone calls from radio and television producers seeking a talking head to confirm the wildly differing -- and usually wrong -- conclusions they've jumped to about what those figures say about the evolving nature of family life in America.
In April, for example, Census officials announced that 56 percent of American children were living in "traditional" nuclear families in 1996, up from 51 percent in 1991. Several prime time television shows excitedly reported this "good news" about the American family, and I heard one radio commentator declare that young couples were finally rejecting the "divorce culture" of their parents' generation.
But this supposedly dramatic reversal of a 30-year trend was based on a peculiarly narrow definition of a traditional family: a two-parent household with children under 18 and no other relatives in the home. If a grandchild, grandparent or other relative were living in the house, the family was "nontraditional." (There's an obvious irony here, given that nothing is quite so traditional as an extended nuclear family that includes a grandparent!)
Evidently, the definition itself was largely responsible for this "trend": Enough such relatives moved to separate households during the first half of the 1990s to increase the proportion of "traditional families," even though the percentage of children living with both biological parents had stayed steady at about 62 percent, and the percentage of married couples had continued its 30-year slide. In other words, the initial reports of a resurgence in traditional families were the result of wishful thinking and a misunderstanding of the terms being used by the census.
But hope springs eternal among talk show producers desperate for a new angle. In mid-May, expanding on its earlier study, the Census Bureau reported that the absolute numbers of married couples with children at home had grown in 2000 after falling in two previous head counts (although the proportion of such families in the total population was still shrinking). TV producers jumped on the story, apparently ready to trumpet the return of the "Ozzie and Harriet" family of the '50s. I soon heard from several talk show hosts in the West who, state-by-state printouts in hand, were agog about the exceptionally large increase of such families in their regions. They wanted me to find them an expert to comment on the heartening return to traditional values.
Their enthusiasm dimmed, however, when I told them that this regional increase in married-couple households with children was due largely to the well-reported influx of Asian and Hispanic immigrants. Their interest evaporated entirely when I reminded them that, as immigrants assimilate, their family patterns tend to match those of the preexisting population.
A week later, the Census Bureau reported that the number of unmarried women with children had increased by 25 percent, dwarfing the 7 percent growth in married-couple families. This time we moved into the "bad news" cycle: Media pundits called to confirm their worst fears, looking for more figures to prove that the explosion of single motherhood was creating an ever-deepening social and cultural crisis in the land.
In fact, most of the growth in "single" motherhood during the 1990s was due to an increase in births to women who, while not married, were living with the child's father. So, much of the recent increase in single motherhood simply reflected the 71 percent increase in cohabitation between 1990 and 2000. But the fact that many "single mother" families actually had fathers present didn't faze the talk show hosts who called seeking confirmation that the sky was falling because of the "collapse of marriage." This time around, they weren't the least bit interested in any good news -- such as the figures, also released in May, that showed a 20 percent drop in births to teenagers over the decade.
Then, last month, the Census Bureau reported that the number of households headed by single fathers had increased fivefold, from 393,000 in 1970 to 2 million in 2000. I got two calls from TV producers that day, each rushing to air a show on this new trend. One asked me to explain how this reflected the increasing equality of men and women in their commitment to parenting, while the other wanted someone to tell her viewers why it represented a backlash against working mothers, who were obviously losing custody to unwed and divorced fathers.
Both producers were crushed when I told them our researchers couldn't confirm either claim, and that we have no way of even knowing how many of these so-called single fathers are in fact living with the mother of their child outside of marriage, and how many are divorced dads who simply happened to have their children with them for the weekend on the day they filled out the form. When I called a Census Bureau researcher to see if he could help straighten this out, he said my guess was as good as his.
It's not that the census researchers are doing a bad job. The problem is that they're asked to compress America's increasingly fluid family arrangements into one-dimensional categories that were established at a time when most single-parent households were created by death rather than by divorce, and when most people made things easy for data collectors by lying rather than admitting to "living in sin."
People's new candor about their lifestyles, combined with the undeniable changes in family arrangements that have occurred over the past 40 years, makes it increasingly hard to capture new family realities in old census categories. And using such categories to talk about families has consequences.
Labeling people single parents, for example, when they may in fact be co-parenting -- either with an unmarried other parent in the home or with an ex-spouse in a joint custody situation -- stigmatizes their children as the products of "single parenthood" and makes the uncounted parent invisible to society. This can lead teachers, school officials, neighbors and other family members to exclude the uncounted parents from activities and interactions into which they might otherwise be drawn. In fear of such marginalization, some separated parents find it hard to agree on a custody arrangement that's in the best interests of the child, because each wants to be the socially recognized parent.
In the past, many "intact" families had fathers who were AWOL from their children's lives. Today, conversely, many "broken" families have fathers who remain active parents. Harvard fellow Constance Ahrons, who has conducted a 20-year study of post-divorce families and their children, has certainly seen plenty of cases where the non-resident parent, usually the father, stops doing any parenting. But she has found many instances where nonresidential fathers became more active in their children's lives after divorce than they were during the marriage. These men need to be recognized for their support, rather than relegated to the same state of non-being as the deadbeat dad.
It's not only parents who are marginalized by outdated household categories. When I speak on work-family issues to audiences around the country, some of the biggest complaints I hear come from individuals who are described by the census as living in "non-family households." They resent the fact that their family responsibilities literally don't "count," either for society or for their employers. There is no category, for instance, for individuals who spend several days a week caring for an aging parent in the parent's separate residence. Yet one in four households in America today is providing substantial time and care to an aging relative, and more than half of all households say they expect to do so within the next 10 years.
It's time for our discussion of family trends to better reflect the complexities of today's family commitments. Perhaps, as Larry McCallum, a therapist who directs the family life program at Augustana College in Rock Island, Ill., suggests, we should do for parents what we have begun to do with racial categories in the census -- provide several alternative ways for people to express their overlapping identities. At the very least, we need to drop the idea that we can predict how a family functions solely by its form.
The place where we keep our clothes isn't always the only place where we keep our commitments.
Stephanie Coontz, professor of history at Evergreen State College in Olympia, Wash., is the national co-chair of the Council on Contemporary Families and author of "The Way We Never Were: American Families and the Nostalgia Trap" (Basic Books).