Are two lesbian parents better than a mom and dad? Logical and methodological flaws in recent studies affirming the superiority of lesbian parenthood.

• Author: Schumm, Walter R.

Biblarz and Stacey cited Satinover for his claim that every child has a need "for both a mother and a father." (1) Later they described such a claim as an "entrenched conviction" that "inflames culture wars over single motherhood, divorce, gay marriage, and gay parenting." (2) They certainly are not alone in questioning such convictions. Recently, numerous scholars have begun to argue that children do not really need a father per se. Peggy Drexler and Linden Gross quoted Dr. Michael Lamb, "It's become clear that the absence of a male figure is really not important," (3) reflecting an apparent scholarly consensus that fathers are no longer necessary for average, much less optimum, child development. (4) Mallon has stated that research is unequivocal in concluding that not even one study has ever found even one disadvantage for children of lesbian or gay parents. (5) In the most recent decade, in an article on lesbian, gay, bisexual, and transgender ("LGBT") families, Biblarz and Savci stated that lesbian mothers "tended to equal or surpass heterosexual married couples on time spent with children, parenting skill, and warmth and affection." (6) While admitting that research was virtually nonexistent with respect to comparisons of married LGBT parents and heterosexual parents (7) or between gay fathers and heterosexual fathers, (8) Biblarz and Stacey reviewed considerable research literature on gender and parenting, drawing several conclusions, as listed below.

(1) "Several studies in Table 1 found that female parents scored higher than heterosexual men on parenting awareness skills and developed warmer, closer, more communicative relationships to their children." (9)

(2) "Two mothers tended to play with their children more and to discipline them less than married heterosexual parents. They were less likely to employ corporal punishment, to set strict limits on their children, or try to elicit social (and gender) conformity." (10)

(3) "In other words, two women who chose to become parents together seemed to provide a double dose of a middle-class 'feminine' approach to parenting." (11)

(4) "We speculate that a double dose of feminine socialization, coupled with discrimination, can lead Heather's two mommies to be among the best, but also somewhat less durable, coparenting couples." (12)

(5) "Research on planned lesbian parenting demonstrates that the impact of this form of 'radical fatherlessness' on children is far from radical, not always fatherless, and arguably more beneficial than not." (13)

(6) "Research to date, however, does not support [the claim that children need both a father and a mother]." (14)

(7) "Our review of research closest to this design suggests that strengths typically associated with mother-father families appear at least to the same degree in families with two women parents." (15)

(8) "In fact, based strictly on the published science, one could argue that two women parent better on average than a woman and a man, or at least than a woman and man with a traditional division of family labor. Lesbian coparents seem to outperform comparable married heterosexual, biological parents on several measures, even while being denied the substantial privileges of marriage." (16)

(9) "Every family form provides distinct advantages and risks for children. Married heterosexual parents confer social legitimacy and relative privilege but often with less paternal involvement. Comothers typically bestow a double dose of caretaking, communication, and intimacy." (17)

(10) "At this point no research supports the widely held conviction that the gender of parents matters for child well-being." (18)

Biblarz and Stacey were willing to admit that two parents in a "low-conflict relationship" probably are better for most children than having only one parent. (19) There are several limitations of the research that are easily overlooked when scholars claim to have found a "consensus" that parental gender does not matter. These limitations, which I will discuss in turn, include: (1) value biases in social science, leading to the marginalization of research or research limitations that do not "fit" the prevailing diversity paradigms; (2) marginalization of methodological problems with null hypothesis testing; (3) theoretical limitations; (4) marginalization of adverse aspects of same-sex parental relationships; and (5) marginalization of adverse aspects of child outcomes associated with same-sex parenting.

My sense is that the root cause of the last four issues is the first issue. Many social scientists are convinced that making distinctions among different family forms is often incorrect and may be generally harmful to those perceived to be minorities. For example, Allen, Fine, and Demo acknowledged that many family scholars believe that "all types of families and ways that families achieve their goals and adjust to their environments are equally adaptive" while they themselves believed that "there is no universally applicable standard type of family or way for families to function." (20) Furthermore, Lerner, Sparks, and McCubbin were concerned that policies that were effective for say, majority families, might be "irrelevant, poorly suited, or even damaging" to minority families. (21) Consequently, the elimination of distinctions could be expected to be more helpful to such families than it might be harmful to majority families, who are seen as unfairly "privileged." (22) Demo et al. commented upon such privilege when they stated, "From a feminist perspective, we assume that power and privilege are systematically distributed inequitably by gender and generation ...." (23) This mindset can create a situation in which obtaining results that do not support distinctions can be seen as inherently helpful even if the theory and/or methodology are weak, whereas obtaining results that would support distinctions might be harmful even if the methodology were strong. Thus, getting the "right" outcome is of far more importance than using more scientifically-sound theoretical or methodological approaches. Again, Demo et al. argued that knowledge should not exist for its own sake but "must be applied to matters of social justice." (24) Of course one might say, "where's the evidence for such bias?"

1. EVIDENCE OF BIAS

   First, Stacey and Biblarz cited Wardle's review in which he claimed there was bias in the field of social science: "an ideological bias favoring gay rights that has compromised most research in this field and the liberal judicial and policy decisions it has informed." (25) Second, while disagreeing with Wardle's views, Stacey and Biblarz themselves recognized the presence of bias in social science, saying:

   We agree, however, that ideological pressures constrain intellectual development in this field .... We wish to acknowledge that the political stakes of this body of research are so high that the ideological "family values" of scholars play a greater part than usual in how they design, conduct, and interpret their studies. (26) They go on to acknowledge that "Wardle ... is correct that contemporary scholarship on the effects of parental sexual orientation on children's development is rarely critical of lesbigay parenthood." (27)

   About the same time of Stacey and Biblarz's review, Redding criticized the lack of sociopolitical diversity in the social sciences, citing the area of lesbigay parenting as a prime example of such bias. (28) Redding cited Tanford, who noted the implications of such bias for the judiciary--"judges may believe that the results of empirical research are unreliable, because they have been distorted by the scientists' liberal values." (29) Erich, Leung, and Kindle also indicated the existence of bias when they said that "[s]ocial justice agendas may have distorted interpretations of research findings." (30) Yet, at the same time, they stated that

   [d]iscrimination against gay and lesbian individuals manifests itself in legislative, social, and institutional obstacles that impair their candidacy to become adoptive parents. This is most dearly evident in Florida where Federal District Judge James Lawrence King denied a gay adult's petition to adopt his foster children on the basis of his sexual orientation. (31) To accuse a federal judge by name (even his middle name!) of discrimination, in my opinion, reflects a certain bias, in itself, by those scholars. More recently, I analyzed over fifty reviews of the literature on lesbigay parenting ill terms of dissertation citations and found that the lower the quality of methodology used, the more likely the dissertations were to be cited in the reviews. (32)

   Later, I turned to a natural experiment in which the same authors from the same university had published three articles in the same time frame using the same sample of lesbian parents, even in the same journal. (33) However, the arguably best article methodologically reported adverse information about lesbian parenting while the other two articles reported favorable information. The latter two articles had been cited sixty-five times, compared to only twice for the former article. (34) Redding acknowledged that the research of Sarantakos had been overlooked but argued that its reliance upon teacher reports was a major methodological issue, despite its relatively large sample size of children. (35) Notably, other scholars have used teacher reports and not found them inherently biased; parent reports are routinely used, despite their obvious potential for self-presentation bias. Later, I will mention the widely overlooked research of Sirota. Thus, the allegation of bias in social science is not a matter of subjective judgment alone, but also a matter of empirical results. It is my belief that this bias is associated with a number of problems in lesbigay parenting research.

2. METHODOLOGICAL PROBLEMS

   Cohen's d, a parameter derived from the difference between the average scores of two groups...