Interesting study done by the Journal of Marriage and Family. Scholars, at USC and New York University, looked at a range of existing studies, including research on gay and lesbian parents, finding that it's ideal if a child is raised by two parents who are "responsible, committed, stable," but that the gender doesn't cause radical differences.
There's also a table at the end of the article that I couldn't replicate its structure here, so do check out the link at the top if you're interested.
http://www3.interscience.wiley.com/cgi-bin/fulltext/123248173/HTMLSTART?CRETRY=1&SRETRY=0
How Does the Gender of Parents Matter?
Timothy J. Biblarz 1 Judith Stacey 2
1 University of Southern California
2 New York University *
Correspondence to Department of Sociology, University of Southern California, Los Angeles, CA 90089-2539 (biblarz@usc.edu).
* Department of Sociology, New York University, 295 Lafayette St., 4th floor, New York, NY 10012.
Copyright Copyright © National Council on Family Relations, 2010
KEYWORDS
bisexual • development or outcomes • family structure • fathering • gay • gender • lesbian • parenting and parenthood • transgender
ABSTRACT
Abstract Method Results Discussion Note References
Claims that children need both a mother and father presume that women and men parent differently in ways crucial to development but generally rely on studies that conflate gender with other family structure variables. We analyze findings from studies with designs that mitigate these problems by comparing 2-parent families with same or different sex coparents and single-mother with single-father families. Strengths typically associated with married mother-father families appear to the same extent in families with 2 mothers and potentially in those with 2 fathers. Average differences favor women over men, but parenting skills are not dichotomous or exclusive. The gender of parents correlates in novel ways with parent-child relationships but has minor significance for children's psychological adjustment and social success.
DIGITAL OBJECT IDENTIFIER (DOI)
10.1111/j.1741-3737.2009.00678.x About DOI
Article Text
Fathers and mothers differ, just as males and females differ.
—David Popenoe
We know the statistics—that children who grow up without a father are five times more likely to live in poverty and commit crime; nine times more likely to drop out of schools and twenty times more likely to end up in prison. They are more likely to have behavioral problems, or run away from home, or become teenage parents themselves.
—Barack Obama
In 1999 American Psychologist unleashed a public furor when it published an article that challenged a popular discourse on the dangers of fatherlessness. "Deconstructing the Essential Father" (Silverstein & Auerbach, 1999) contended that successful parenting is not gender specific and that children do not need fathers, or mothers either, for that matter. Rather, any gender configuration of adults could parent well. The implication that fathers were expendable incited an uproar. Wade Horn (1999), soon to become Secretary for Children and Families at the Department of Health and Human Services, labeled the article "Lunacy 101: Questioning the Need for Fathers," and other critics were more vitriolic (e.g., Jacoby, 1999).
Complex scholarly questions about the significance of parental gender were lost in the firestorm, and the view that Silverstein and Auerbach (1999) challenged continues to dominate public policy. Thus, the 2006 New York Court of Appeals ruling against same-sex marriage found that "the Legislature could rationally believe that it is better, other things being equal, for children to grow up with both a mother and a father. Intuition and experience suggest that a child benefits from having before his or her eyes, every day, living models of what both a man and a woman are like" (Justice Robert Smith in Hernandez v. Robles, 2006). Presidents from both political parties concur. Former President Bush defended Florida's ban on gay adoption rights contending that "studies have shown that the ideal is where a child is raised in a married family with a man and a woman" (quoted in Bumiller, Sanger, & Stevenson, 2005). President Obama endorsed stereotypical views about fathers in 2008: "Of all the rocks upon which we build our lives, we are reminded today that family is the most important. And we are called to recognize and honor how critical every father is to that foundation. They are teachers and coaches. They are mentors and role models. They are examples of success and the men who constantly push us toward it" (New York Times, 2008).
The argument that children need both a mother and father presumes that mothering and fathering involve gender-exclusive capacities. The "essential father" is a disciplinarian, problem solver, and playmate who provides crucially masculine parenting. Boys need fathers, proponents claim (Blankenhorn, 1995; Popenoe, 1996; Wilson, 2002), to develop appropriate masculine identity and to inhibit antisocial behaviors like violence, criminality, and substance abuse; in contrast, fathers foster heterosexual femininity in daughters and help deter promiscuity, teen pregnancy, and welfare dependency. Dad "is the grinding stone on which his son sharpens his emerging masculinity and the appreciative audience to which his daughter plays out her femininity" (Pruett, 2000, p. 87). Mothers provide nurturance, security, and caretaking.
The belief that research supports these convictions remains widespread. At the risk of inviting charges of "Lunacy 201," we wish to revive conversation among scholars about research on gender differences in parenting and child development. Researchers agree that, on average, women and men parent somewhat differently, but they do not agree on the sources, fixity, or consequences of these differences. Beyond lactation, are there exclusively female or male parenting abilities? Does female-only and male-only parenting differ? Does fatherless or motherless parenting create particular difficulties or opportunities for children, and are these the same for girls and boys? Research provides few unambiguous answers.
We undertake a careful review of relevant research to assess what it can contribute to understanding how the gender of parents matters. There is no clearly identified body of research, however, to which we can easily turn. An unusually diffuse array of literatures bears indirectly on these questions: studies of primary caretaker married fathers; egalitarian coparenting heterosexual couples; single mothers or fathers after a death, divorce, or desertion; heterosexual single mothers by accident or choice; lesbian mothers and gay fathers after heterosexual divorce; planned lesbian motherhood through donor insemination (DI) or adoption; planned gay fatherhood through adoption, surrogacy, or coparenting with women. Thus, to make headway on these issues we must extract scattered kernels of empirical wheat from masses of extraneous chaff. We begin by identifying genres of research from which claims about gender differences have been drawn and show that they were not designed to address these questions, nor are they capable of doing so. Next we analyze studies that do bear on these questions. Finally, we suggest a research strategy to better address intriguing unresolved questions.
Misleading Representations of Research
Conventional understanding of gender differences in parenting derives primarily from studies of married mothers and fathers. The gist of this research is unsurprising: Most married wives exceed their husbands relatively and absolutely on time spent in child care and domestic work and on most types of interaction with their children (e.g., Hall, Walker, & Acock, 1995; Hawkins, Amato, & King, 2006; Yeung, Sanberg, Davis-Kean, & Hofferth, 2001). Fathering among married couples involves more breadwinning, stereotypical masculine tasks, and play with children (e.g., Hawkins et al.). Married fathers generally spend more time with sons than daughters (Harris, Furstenberg, & Marmer, 1998; Marsiglio, 1991) and express greater interest in children's gender conformity (e.g., Pruett, 2000). These patterns have been softening over time (e.g., Bianchi, 2000; Sandberg & Hofferth, 2001), and parenting differences between families often exceed gender differences within families (e.g., Lytton & Romney, 1991). Despite stereotypes about fathers as disciplinarians, mothers physically punish children more often than fathers do because they spend much more time caring for children. Relative to the amount of time spent caring for children, however, fathers commit more physical and sexual abuse (Sedlack & Broadhurst, 1996; Straus, 2001).
The research on married couples largely reinforces what we already know—that women are more likely than men to commit to children, relationships, and homemaking. What we do not know is whether these average differences derive from gender per se, from heterosexual gender, or from other factors. Because these studies do not include single-sex parent families, they do not isolate the effects of parental gender, as public officials, advocates, and even some social scientists infer. Our two epigraphs about parental gender differences and the distinctive role of fathers rest on a similar analytical error. They draw inappropriately from studies of divorce and single mothering and conflate findings on five distinct variables that interact in complex ways—number of parents, gender, sexual identity, marital status, and biogenetic relationship to children. In a typical and weighty example of this error, an expert witness for the state of Washington drew on an "extensive review of the published research on the need that every child has for both a mother and a father" (Satinover, 2004, p. 3) to oppose same-sex marriage. The psychiatrist testified that research confirmed that "children not raised by their own married mother and father are subject to increased risk of disadvantage and harm" (p. 9). He attested that "quantifiable deficits occur in literally every area of development" (p. 10) in children who "do not live with both their biological mother and father" (p. 9). Almost all of the studies Satinover cited, however, compared single-mother with married-parent families. None compared children raised by two female parents with those raised by two males or by one male and one female. None compared primary-caretaker fathers and mothers or children adopted by single men and women.
Reputable social scientists have issued similar misinterpretations of research. Prominent scholars signed a report, "Why Marriage Matters: Twenty-One Conclusions from the Social Sciences" (Glenn, Nock, & Waite, 2002). Echoing errors in Satinover's (2004) affidavit, the report analyzed research in which marital status, rather than gender, was the primary parental variable, but it conflated marriage with different-sex couples. To support its claim that "a child who is not living with his or her own two married parents is at greater risk of child abuse" (p. 17), the report cited studies of children who live alone with single mothers, in stepfamilies, or with their mother's boyfriends. It ignored research on lesbian and gay parenthood.
Because access to legal same-sex marriage is so new and rare, we do not yet have research that compares the children of married same-sex and different-sex couples. Even so, scholars have achieved a rare degree of consensus that unmarried lesbian parents are raising children who develop at least as well as their counterparts with married heterosexual parents (e.g., American Academy of Pediatrics, 2002; Stacey & Biblarz, 2001; Tasker, 2005).
Additionally, most family researchers agree that, all other things being equal, two parents (in a low-conflict relationship) generally provide more material and emotional resources to children than one parent (e.g., Amato, 2005; McLanahan & Sandefur 1994). This research, however, did not compare children in married-couple homes with children raised by same-sex couples. Moreover, too often it compared children in married-couple and single-parent families, thereby confounding the effects of number of parents with those of their marital history. None of this research investigated whether the gender of the absent parent was responsible for different child outcomes in single- versus two-parent families. Nonetheless, consequential policy decisions rest on the view that "[f]ew propositions have more empirical support in the social sciences than this one: Compared to all other family forms, families headed by married, biological parents are best for children" (Popenoe, quoted in Center for Marriage and Families, 2006, p. 1).
Research on Parental Sex Differences and Child Outcomes
Well, then, "where's the wheat?" What research literature contributes fruitful knowledge kernels about sex differences in parenting? Surprisingly few studies examined how the gender, as distinct from the number, marital status, sexual orientation, or biogenetic relationship of parents, affects children. No study attempted to isolate the variable of parental gender by holding constant these other factors. No research compared planned parenting by couples composed of women only, men only, or a woman and a man. No studies compared single-parent adoptions by women and men. Consequently, the gender differences question requires an indirect, inductive approach. Two bodies of research avoid some of these problems. Some studies explored parenting differences between biological, presumptively heterosexual, single fathers and mothers. Here the gender of parent varied, but the number, marital status, and genetic relationship to children did not. A related genre of studies explored differences in parenting between married wives and their husbands but did not compare single-sex with different-sex coparent families. The second, most germane body of research compared planned lesbian coparent families with heterosexual married-parent families. Although this research investigated the impact of sexual identity on parenting and children, it offers data more relevant to parental gender, particularly "fatherlessness," because it compared families that did and did not include male parents. A new promising avenue of research compared families headed by lesbians to those headed by gay male parents (Johnson & O'Connor, 2002).
Although these research designs help to disentangle the gender of parents from the confounding variables of number and marital status, they also have limitations. Studies that compared single (heterosexual) mothers and fathers rarely controlled effects of diverse routes to single parenthood—chosen or accidental or via the loss of a coparent through death, desertion, or a divorce in which child custody was contested or granted willingly. Studies that compared lesbian comothers (or gay cofathers) with heterosexual coparents, on the other hand, rarely could control for marital status or biological relatedness of both parents, and they could not readily distinguish the impact of gender from sexual identity. Moreover, these studies were conducted in different states and nations with distinct and changing sociocultural and legal contexts for parenting, such as the Netherlands, where same-sex marriage was legalized in 2001.
Method
Abstract Method Results Discussion Note References
Despite these limitations, a careful analysis of these two bodies of research sheds light on ways in which growing up with both a mother and a father does and does not matter for children and our society. We examined all pertinent studies published from 1990 onward located through database searches of PsychINFO, Sociological Abstracts, JSTOR, and ProQuest; by pursuing references in published studies; and through personal contacts with family researchers. To be included, studies had to report findings on parenting or child outcomes, or both; statistically assess significance of differences between groups; and compare families with the same number of residential parents but different configurations of male and female parents. (An appendix listing key methodological features of each study—publication type, target population, type and size of samples and comparison groups, response rate, primary dependent variables, method of data collection, and statistical approach to the data—is available on the Journal of Marriage and Family website). We were concerned about treating findings from studies that were somewhat weaker methodologically as equal to those from relatively stronger studies, so we experimented with rating studies on methodological dimensions. Because studies of the same dependent variables tended to log similar differences between groups even as their methodological strength varied, however, we include findings from all 33 studies of two-parent families and 48 studies of single-parent families. Thirty of the two-parent studies compared lesbian to heterosexual coparents, one compared gay male to heterosexual coparents, and two compared lesbian to gay male coparents.
Most studies of heterosexual single-parent families drew on surveys with national probability samples and self-administered questionnaires. In contrast, comparative studies of lesbian and heterosexual coparent families typically employed snowball samples or drew from fertility clinic rosters of couples who employed DI. These usually included in-depth interviews and observations and employed psychometric instruments that assess the well-being of family members and the quality of their relationships. Although researchers who examine the impact of parental sex orientation on children reported few significant differences in child outcomes between children raised by heterosexual and lesbian couples (e.g., Tasker, 2005; Telingator & Patterson, 2008), the overwhelming public consensus is that children raised by both a mother and father develop more successfully.
To vet all evidence that speaks to this counter-consensus, we focus our analysis on the comparatively rare statistically significant findings of difference between families with same- and different-sex parents or female versus male parents. Tables 1 and 2 cross-classify findings of difference from the 81 studies by whether they involved two-parent (Table 1) or single-parent families (Table 2) and whether they pertained to parents (top panel) or children (bottom panel). The findings of difference we show are general summaries of numerous individual tests within and across studies. Even so, for every finding of significant difference in studies of same- and different-sex coparent families, there were roughly four or more findings of no significant difference that we do not display. Our focus on findings of difference increases the risk of Type I error. Because 5 of every 100 differences may occur by chance at p≤ .05, as many as 20% (5/25) of the findings shown might be noise. On the other hand, some of the findings of no differences may miss real differences (Type II error) because some studies use levels of significance that may be too restrictive for their very small samples. Risking Type I error is less of a problem in studies that compare single mothers with single fathers because nearly half of these tests for difference were statistically significant.
There's also a table at the end of the article that I couldn't replicate its structure here, so do check out the link at the top if you're interested.
