Last months Senate hearing on a bill to repeal the federal Defense of Marriage Act (DOMA) featured a clash between Sen. Al Franken (D-MN) and one of the witnesses defending DOMA, Tom Minnery of Focus on the Family.

Minnerys testimony referred to the social science evidence showing children do best when raised by their own mother and father. He referred to one such study in his prepared testimony this way:

In fact, the U.S. Department of Health and Human Services explains in its new and exhaustive report, Family Structure and Childrens Health in the United States: Findings from the National Health Interview Survey, 2001-2007, that children living with their own married biological or adoptive mothers and fathers were generally healthier and happier, had better access to health care, less likely to suffer mild or severe emotional problems, did better in school, were protected from physical, emotional and sexual abuse and almost never live in poverty, compared with children in any other family form.

Franken, however, triumphantly noted that in fact, these superior outcomes were associated with nuclear families, defined as one or more children living with two parents who are married to one another and are each biological or adoptive parents to all children in the family. Since the definition made no mention of the gender of the married parents, he concluded that nuclear families could be headed by married homosexual couples, too.

FRCs Tony Perkins, however, noted in his Washington Update that Franken seemed to be forgetting the very law that the hearing was about:

DOMA says, "In determining the meaning of . . . any ruling, regulation, or interpretation of the various administrative bureaus and agencies of the United States, the word 'marriage' means only a legal union between one man and one woman as husband and wife." Since this was a federal study published by a federal agency based on a federal survey conducted by federal (Census Bureau) employees, its definition of married is bound by DOMA.

I had made the same point in a longer op-ed about this study in February.

Just to be sure, however, I sent an email to the Centers for Disease Control and Prevention (CDC), which had published the study. Unfortunately, they confirmed that they had simply ignored the mandate of DOMA with respect to the definition of marriage.

Here, for the record, is the substantive part of their response:

The National Health Interview Survey (NHIS) is conducted under the authority of the National Center for Health Statistics and obtains annual data on various health characteristics (e.g., health status, chronic conditions, disability, access to health care, etc.) of the US noninstitutionalized population. NHIS data are typically obtained via face-to-face computer-assisted interviews using fixed format questions. All NHIS data are based on self reports by respondents who decide for themselves how they wish to answer each question; the interviewer then enters the response on the computer.

Regarding marital status, all household members aged 14 or older are asked if they are now married, widowed, divorced, separated, never married, or living with a partner. NHIS respondents self-identify whether they are currently married, divorced, living with a partner, etc.; they are not asked questions about the type of union (e.g., civil unions, common-law marriages, etc) or date of divorce, whether either the marriage or divorce is legal according to the state they live in, or how long their union has lasted. If NHIS respondents tell us they are married, we accept that response as is. Other than making sure that the data are consistent with the universe (that is, limited to respondents aged 14 or older), no attempt is made by NCHS staff to correct the data. Moreover, the 2001-2007 NHIS did not contain questions that systematically asked about sexual orientation; gay or lesbian respondents, as well as same-sex couples, are contained in the data but are not identified. As a result, the definitions used in sr10_246 (the report on family structure and children's health) were neutral regarding the gender of parents.

It seems to me that it would be easy enough in the interview process to briefly explain the federal definition of marriage, so that the respondent can reply to the marital status question in a way consistent with the law. Procedures for conducting these interviews should immediately be amended to bring them into conformity with DOMA.

Nevertheless, I will give one round to Franken on this technical point. But the number of nuclear families headed by homosexual couples in this study is likely to be negligible. As Tony Perkins noted in the same Washington Update piece quoted above,

Even if, by chance, the interviewers or authors violated [DOMA], the survey data was collected from 2001 to 2007. During that time (and only from mid-2004 on) there was only one state (Massachusetts) in which homosexual couples could marry.

Furthermore, even married homosexual couples who are raising children are unlikely to fit the definition of a nuclear family. Remember, a nuclear family requires that the married parents are each biological or adoptive parents to all children in the family (emphasis added). Obviously, it is biologically impossible for same-sex partners to each be biological parents of the same children.

Only if a married same-sex couple had jointly adopted all the children in their household would they clearly fit even this studys lawless definition of a nuclear family. However, this is not the most common type of homosexual parenting situation. It is much more common for homosexual parents to bring to their relationships their own biological childrenconceived in previous heterosexual relationships.

Advocates for homosexual parents (and for homosexual marriage) are fond of arguing that children do not need a mom and a dadall they need is two loving parents. But this HHS study, while unfortunately not consistent with federal law in its definition of married parents, still offers no support for that argument.

Households featuring same-sex couples raising children are much more likely to fit one of the other non-nuclear household typesall of which, apart from single-parent families, feature at least two adult caretakers. These include unmarried biological or adoptive, blended, cohabiting, extended, or other families.

All of these family types had outcomes inferior to those of the nuclear family.