Support First Things by turning your adblocker off or by making a  donation. Thanks!

As if it didn’t have enough to fret about, the two-parent American family got taken for quite a ride in the past year. First, in April the Census Bureau dramatically reported that the “nuclear family” was “rebounding.” The page-one story in USA Today announced: “The traditional nuclear family”a married mom and dad living with their biological children”is making a comeback, according to a Census report released today. The proportion of the nation’s children living with both biological parents jumped from 51 percent in 1991 to 56 percent . . . in 1996.” On ABC World News Tonight , Peter Jennings declared: “The Census Bureau said today that the number of children who live with both their parents increased during the 1990s.” Scores of news organizations around the country reported the same happy story.

But in May, the story reversed itself. Journalists across the country began to report that “nuclear families” now constitute less than 25 percent of all U.S. households. An editorial in the New York Post announced that “the American nuclear family” was now “up there with the Pacific salmon as an endangered species.” A Newsweek cover story on unmarried mothers (“the new faces of America’s family album”) explained at length how the “traditional family” was “fading fast.” Dr. James Dobson, the radio personality and president of Focus on the Family, an influential Christian pro-family organization, said that the “alarming” Census Bureau figures revealed “just how dire the situation has become,” as “the family is unraveling at a faster pace than ever.” The New York Times took a sunnier view. Following up on its page-one news story (“Number of Nuclear Families Drops as 1-Parent Families Rise”), the Times editorial board urged its readers not to worry about the decline, since “the nuclear family is not the only kind of family or even the only healthy kind of family.”

Well, now. Is it possible for the nuclear family to be simultaneously “making a comeback” and “fading fast”? Of course not. So which is true? Amazingly enough, the answer is neither . What is actually happening to U.S. family structure is quite different from the news conveyed by either cycle of stories.

While the Census Bureau has been quick to blame the media for the confusion, the fault lies primarily with the Bureau. For mysterious reasons, the Census Bureau chose in the fine print of its April report to define a “traditional nuclear family” as a household consisting of two biological parents, their minor children, and no one else . That is, a household that includes grandparents is not “traditional.” A household that includes boarders, or a foster child, is not “traditional.” Moreover, during the 1990s, for reasons that have almost nothing to do with the core issue”which is how many U.S. children are growing up in households with two married parents”three-generation and large or complex households with children declined slightly as a proportion of all households with children.

That little curio of a demographic fact”that tangent of a tangent”was the entire basis of the Census Bureau’s April “rebound” report. No evidence presented in that report justifies the assertion that the proportion of children living with both biological parents rose during the early- and mid-1990s. And to add injury to insult, the obscurantist definition of “traditional” means that the Census Bureau, while allegedly describing a nuclear family “rebound,” actually underreported the proportion of U.S. children living with two married parents. The Census Bureau report puts the figure at 56 percent for 1996. But my own research, subsequently confirmed by other researchers and the Census Bureau itself, shows that the actual figure for 1996 is 64 percent.

As for May’s precipitous drop in optimism, it can be traced to further manipulation of definitions on the part of journalists with no helpful guidance and arguably even some complicity from the Census Bureau. This time, using a data table released in mid-May with much fanfare by the Census Bureau, journalists chose to measure married-couple-with-children families not as a proportion of all families (two or more persons living together related by blood, marriage, or adoption), but instead as a proportion of all households . Grandma living on her own is a household. College roommates sharing an apartment are a household. The number of non-family households in the U.S. has been growing steadily for many decades for multiple reasons (including longer life spans and greater affluence), most of which have little to do with the state of marriage and child rearing. Indeed, marrieds-with-children were a distinct minority of all U.S. households even in the 1950s. That’s why most scholars agree that the best way to measure the prevalence of a family phenomenon such as marriage is to place it in the context of family households; throwing in non-family households is like mixing apples and oranges. This piece of confusion was the primary basis for the “decline” stories that received so much attention last May.

This episode is a distressing example of irresponsibility by a public agency charged with collecting and reporting data on how we live. Even at this late date, it is impossible for scholars to get accurate trend-line data from the Census Bureau on the proportion of U.S. children living with their two biological married parents. This past June, nine senior family scholars, led by Norval Glenn of the University of Texas and Linda Waite of the University of Chicago, wrote a public letter to the Census Bureau asking it to disentangle the definitions and report this basic information. The Census Bureau politely declined. For the time being, at least, accurate information about this trend will have to come from elsewhere.

Which brings us to the truly good news. A series of recent reports from independent scholars, plus largely unpublished data from the 2000 Census, all suggest that the trend of family fragmentation that many analysts had assumed to be unstoppable suddenly stopped in its tracks about six years ago.

What we are seeing is not (at least not yet) a “rebound.” But it’s certainly not a “decline.” To be conservative, let’s call it a cessation, a significant pause. But let’s say it more optimistically: after more than three decades of relentless advance, the family structure revolution in the U.S. may be over.

Here are the basic numbers. The proportion of all U.S. families with children under age eighteen that are headed by married couples reached an all-time low in the mid-1990s”about 72.9 percent in 1996 and 72.4 percent in 1997”but has since stabilized. The figure for 2000 is 73 percent. Similarly, the proportion of all U.S. children living in two-parent homes reached an all-time low in the mid-1990s, but since then has stabilized as well. In fact, the proportion of children in two-parent homes increased slightly from 68 percent in 1999 to 69.1 percent in 2000.

Looking only at white, non-Hispanic children, a study by Allan Dupree and Wendell Primus finds that the proportion of these children living with two married parents stopped its downward descent during the late 1990s, and even increased modestly from 1999 to 2000, rising from 77.3 to 78.2 percent. Another study from the Urban Institute finds that, among all U.S. children, the proportion living with their two biological or adoptive parents increased by 1.2 percent from 1997 to 1999, while during the same period the proportion living in stepfamilies (or blended families) decreased by 0.1 percentage points and the proportion living in single-parent homes decreased by two percentage points. (The study finds that in 1999 about 64 percent of all U.S. children lived with their two biological or adoptive parents, while about 25 percent lived with one parent and about 8 percent lived in a step or blended family.) Among low-income children, the decline in the proportion living in single-parent homes was even more pronounced, dropping from 44 percent in 1997 to 41 percent in 1999.

And, perhaps most encouraging, from 1995 to 2000 the proportion of African-American children living in two-parent, married-couple homes rose from 34.8 to 38.9 percent, a significant increase in just five years, representing the clear cessation and even reversal of the long-term shift toward black family fragmentation.

What has caused this shift? No one knows for sure, but we can make some plausible guesses. The roaring economy probably had little or nothing to do with it, since all previous economic booms since 1970 have coincided with growing family fragmentation, not reintegration. On the other hand, federal and state welfare reforms dating from the mid-1990s, which dramatically restructured and in some instances eliminated what had previously been guaranteed economic supports for unmarried mothers, have almost certainly played a role. As the above-cited data suggest, post-1995 family structure changes have been most dramatic among low-income families.

More generally, on the core social question of whether family fragmentation is a bad thing or a not-so-bad thing, a steady shift in popular and (especially) elite opinion took place over the course of the 1990s. Denial and happy talk about the consequences of nuclear family decline became decidedly less widespread; concern and even alarm became much more common. As a society we changed our minds, and as a result we changed some of our laws. And now, it seems, we are beginning to change some of our personal behavior. This is very encouraging news.

It is now clear that those who have long and loudly insisted that nothing can be done to stop the trend of family fragmentation are wrong. Remember all their clichés? We have to be realistic, they opined. The “family diversity” trend is irreversible. We can’t put the toothpaste back in the tube. We shouldn’t fall into the “nostalgia trap.” We can’t go back to “Ozzie and Harriet.” Well, the next time someone tells you that, just smile and show him the new numbers. Positive change in U.S. family structure is not only desirable and possible. It is already occurring. Today our main challenge is no longer to reverse a trend toward disintegration, but to intensify the nascent trend toward reintegration.

David Blankenhorn is president of the Institute for American Values.