Funny Numbers Pervade Census Data

The traditional American family had a terrific April followed by a terrible May. "Nuclear Family Makes Comeback," said the Pittsburgh Post-Gazette on April 13. On May 15, a Post-Gazette headline and story announced the opposite: "Two-Parent Families Less Common." The Los Angeles Times had contradictory headlines, too: "Report: More Children Lived in Nuclear families in'90s," then a month later, "Married With Children Is on the Wane in the U.S."

Thank the Census Bureau for the confusion. It managed to turn out back-to-back reports, one saying that the nuclear family was recovering, the other announcing that the nuclear family is tanking. So the American public was badly misinformed about families at least once, and probably twice.

In April, the bureau released a survey: "Living Arrangements of Children: Fall 1996." "The nuclear family rebounds," said the official press release. It said that the percentage of American children living with their married biological parents had jumped from 51 percent to 56 percent in the years 1991-1996.

But the rise was imaginary. It was based on the peculiar way the census people keep family statistics: a mom and dad living with their biological children don't count as traditional if another person lives in the household, a boarder perhaps, or a relative. With the economic boom, many grandparents who were staying with a married son or daughter found they could afford their own housing. In the eyes of the Census Bureau, each grandparent who moved out created a new nuclear family out of those left behind. Columnist Maggie Gallagher said the correct headline should have been "Married Families Less Likely to Live With Gramps."

In fact, the proportion of children living with their married biological parents remained steady at 62 percent from 1991 to 1996. David Blankenhorn of the Institute for American Values got that number by diving into the bureau's statistics and doing his own arithmetic. When questioned by Gallagher, census official Jason Fields said Blankenhorn's number was correct, though he did not explain why the public should have to figure out this important statistic on its own. "Children are neither more or less likely to live with their own two biological parents" than they were in 1991, he said, acknowledging that no "rebound" had actually occurred.