In nearly every discussion, debate or lawsuit about gay marriage, the talk at some point turns to family values.
Do gay couples make for good parents? Will their children -- whether adopted, conceived with the help of a surrogate or brought in from a pre-existing relationship -- adjust, adapt and succeed in a world dominated by traditional families?
The answers usually depend on who's giving them, and come dressed in anecdotes and colored by bias. But Stanford sociologist Michael Rosenfeld brings something new to the conversation: facts and figures derived from the country's largest data bank -- the U.S. Census.
In a study published this month in the journal Demography, Rosenfeld concludes that children being raised by same-sex couples have nearly the same educational achievement as children raised by married heterosexual couples.
By mining data from the 2000 Census, Rosenfeld was able to figure out the rates at which children in all types of families repeated a grade during elementary or middle school. According to his findings, nearly 7 percent of children raised by heterosexual married couples were held back a year, while about 9.5 percent of children living with adults identifying themselves as same-sex partners repeated a grade.
The difference between the groups pretty much vanishes when taking into account that the heterosexual couples were slightly more educated and wealthier than most gay parents, Rosenfeld said.
"The census data show that having parents who are the same gender is not in itself any disadvantage to children," he said. "Parents' income and education are the biggest indicators of a child's success. Family structure is a minor determinant."
Rosenfeld's findings have been cited by lawyers fighting Proposition 8, the gay marriage ban passed by California voters in 2008. A federal court judge recently overturned the ban, but his ruling is under appeal.
Rosenfeld's study shows that children of gay and married couples had lower grade-repetition rates than their peers raised by opposite-sex unmarried couples and single parents. And all children living in some type of family environment did much better than those living in group housing. Those who were awaiting adoption or placement in a foster home were held back about 34 percent of the time.
"One of the fundamental issues in modern family law that differs from state to state is whether same-sex couples can adopt," Rosenfeld said. "My research makes clear that there's a huge advantage to kids to be out of the care of the state and into the care of any family, even if the family is not perfectly optimal."
Educators, policymakers and social scientists have long known that children left back in school are at greater risk than their peers for not finishing high school and getting into trouble.
Because gays and lesbians make up such a tiny sliver of the American population -- only 1 percent -- it has been difficult for researchers to conduct a representative study of how their children perform in the classroom. And gay marriage opponents have criticized earlier studies for having sample sizes that are too small.
"Sample size is power," Rosenfeld said. "And the census is the biggest sample we have. This study is based on a sample of thousands and thousands of kids."
Most personal decisions about gay marriage are based on gut feelings, religious beliefs and individual experiences. Rosenfeld knows his research isn't going to change the minds of most people opposed to same-sex unions. But he has added new data to the debate that helps debunk assertions -- whether based on a lack of knowledge or some unfounded fear -- that children raised by gay couples cannot thrive.
"Social scientists have an obligation to shed light where they can on issues that are roiling the public," he said. "Sometimes we have to throw up our hands and admit that something is unknowable. But in this case, we could bring some real hard data to bear on an area that was otherwise really in the dark."
- Michael J. Rosenfeld. Nontraditional Families and Childhood Progress Through School. Demography, 2010; 47 (3): 755 DOI: 10.1353/dem.0.0112
Cite This Page: