Study On Homosexual Parents Tops All Previous Research
In a historic study of children raised by homosexual parents, sociologist Mark Regnerus of the University of Texas at Austin has overturned the conventional academic wisdom that such children suffer no disadvantages when compared to children raised by their married mother and father. Just published in the journal Social Science Research,[1] the most careful, rigorous, and methodologically sound study ever conducted on this issue found numerous and significant differences between these groups--with the outcomes for children of homosexuals rated "suboptimal" (Regnerus' word) in almost every category.
The Debate Over Homosexual Parents
In the larger cultural, political, and legal debates over homosexuality, one significant smaller debate has been over homosexual parents. Do children who are raised by homosexual parents or caregivers suffer disadvantages in comparison to children raised in other family structures--particularly children raised by a married mother and father? This question is essential to political and ethical debates over adoption, foster care, and artificial reproductive technology, and it is highly relevant to the raging debate over same-sex "marriage." The argument that "children need a mom and a dad" is central to the defense of marriage as the union of one man and one woman.
Here is how the debate over the optimal family structure for children and the impact of homosexual parents has usually gone:
Pro-family organizations (like Family Research Council) assert, "Social science research shows that children do best when raised by their own biological mother and father who are committed to one another in a life-long marriage." This statement is true, and rests on a large and robust collection of studies.Pro-homosexual activists respond, "Ah, but most of those studies compared children raised by a married couple with those raised by divorced or single parents--not with homosexual parents." (This is also true--in large part because the homosexual population, and especially the population of homosexuals raising children, is so small that it is difficult to obtain a representative sample.)The advocates of homosexual parenting then continue, "Research done specifically on children raised by homosexual parents shows that there are no differences (or no differences that suggest any disadvantage) between them and children raised by heterosexual parents."
Pro-family groups respond with a number of critiques of such studies on homosexual parents. For example, such studies usually have relied on samples that are small and not representative of the population, and they frequently have been conducted by openly homosexual researchers who have an ideological bias on the question being studied. In addition, these studies also usually make comparisons with children raised by divorced or single parents--rather than with children raised by their married, biological mother and father.
In fact, an important article published in tandem with the Regnerus study (by Loren Marks, Louisiana State University) analyzes the 59 previous studies cited in a 2005 policy brief on homosexual parents by the American Psychological Association (APA).[2]
Marks debunks the APA's claim that "[n]ot a single study has found children of lesbian or gay parents to be disadvantaged in any significant respect relative to children of heterosexual parents."
Marks also points out that only four of the 59 studies cited by the APA even met the APA's own standards by "provid[ing] evidence of statistical power." As Marks so carefully documents, "[N]ot one of the 59 studies referenced in the 2005 APA Brief compares a large, random, representative sample of lesbian or gay parents and their children with a large, random, representative sample of married parents and their children."
To summarize, we have been left with large, scientifically strong studies showing children do best with their married mother and father--but which do not make comparisons with homosexual parents or couples; and studies which purportedly show that children of homosexuals do just as well as other children--but which are methodologically weak and thus scientifically inconclusive.
The New Family Structures Study--Restoring the "Gold Standard"
This logjam of dueling studies has been broken by the work that Regnerus has undertaken. Unlike the many large studies previously undertaken on family structure, Regnerus has included specific comparisons with children raised by homosexual parents. Unlike the previous studies on children of homosexual parents, he has put together a representative, population-based sample that is large enough to draw scientifically and statistically valid conclusions. For these reasons, his "New Family Structures Study" (NFSS) deserves to be considered the "gold standard" in this field.
Another improvement Regnerus has made is in his method of collecting data and measuring outcomes for children in various family structures. Some previous studies collected data while the subjects were still children living at home with their parent or parents--making it impossible to know what the effects of the home environment might be once they reach adulthood. Some such studies even relied, in some cases exclusively, on the self-report of the parent. This raised a serious question of "self-presentation bias"--the tendency of the parent to give answers that will make herself and her child look good.
Regnerus, on the other hand, has surveyed young adults, ages 18 to 39, and asked them about their experiences growing up (and their life circumstances in the present). While these reports are not entirely objective, they are likely to be more reliable than parental self-reports, and allow evaluation of long-term impacts.
The study collected information from its subjects on forty different outcomes. They fall into three groups:
To summarize, we have been left with large, scientifically strong studies showing children do best with their married mother and father--but which do not make comparisons with homosexual parents or couples; and studies which purportedly show that children of homosexuals do just as well as other children--but which are methodologically weak and thus scientifically inconclusive.
The New Family Structures Study--Restoring the "Gold Standard"
This logjam of dueling studies has been broken by the work that Regnerus has undertaken. Unlike the many large studies previously undertaken on family structure, Regnerus has included specific comparisons with children raised by homosexual parents. Unlike the previous studies on children of homosexual parents, he has put together a representative, population-based sample that is large enough to draw scientifically and statistically valid conclusions. For these reasons, his "New Family Structures Study" (NFSS) deserves to be considered the "gold standard" in this field.
Another improvement Regnerus has made is in his method of collecting data and measuring outcomes for children in various family structures. Some previous studies collected data while the subjects were still children living at home with their parent or parents--making it impossible to know what the effects of the home environment might be once they reach adulthood. Some such studies even relied, in some cases exclusively, on the self-report of the parent. This raised a serious question of "self-presentation bias"--the tendency of the parent to give answers that will make herself and her child look good.
Regnerus, on the other hand, has surveyed young adults, ages 18 to 39, and asked them about their experiences growing up (and their life circumstances in the present). While these reports are not entirely objective, they are likely to be more reliable than parental self-reports, and allow evaluation of long-term impacts.
The study collected information from its subjects on forty different outcomes. They fall into three groups:
Some are essentially yes-or-no questions: are you currently married, are you currently unemployed, have you thought recently about suicide?
Other questions asked respondents to place themselves on a scale--for example, of educational attainment, happiness or depression, and household income.
Finally, "event-count" outcomes involve reporting the frequency of certain experiences--e.g., smoking marijuana or being arrested--and the number of sex partners.
Nearly 15,000 people were "screened" for potential participation in the study; in the end almost 3,000, a representative sample, actually completed the survey questionnaire. Of these, 175 reported that their mother had a same-sex romantic relationship while they were growing up, and 73 said the same about their father. These are numbers just large enough to make some statistically robust conclusions in comparing different family structures.
What the Study Found
The study looked at 40 different outcomes, but reported data for children with "lesbian mothers" and those with "gay fathers" separately. Therefore, there actually were 80 outcome measures that could be said to compare children with "homosexual parents" to those from other family structures. When compared with outcomes for children raised by an "intact biological family" (with a married, biological mother and father), the children of homosexuals did worse (or, in the case of their own sexual orientation, were more likely to deviate from the societal norm) on 77 out of 80 outcome measures.(The only exceptions: children of "gay fathers" were more likely to vote; children of lesbians used alcohol less frequently; and children of "gay fathers" used alcohol at the same rate as those in intact biological families).
Of course, anyone who has had a college course in statistics knows that when a survey shows there are differences between two groups, it is important to test whether that finding is "statistically significant." This is because it is always possible, by chance, that a sample may not accurately reflect the overall population on a particular point. However, through statistical analysis researchers can calculate the likelihood of this, and when they have a high level of confidence that a difference identified in the survey represents an actual difference in the national population, we say that finding is "statistically significant." (This does not mean the other findings are unimportant--just that we cannot have as high a level of confidence in them.)
Regnerus has analyzed his findings, and their statistical significance, in two ways--first by a simple and direct comparison between what is reported by the children of homosexual parents and the children of "intact biological families" ("IBFs"), and second by "controlling" for a variety of other characteristics. "Controlling for income," for example, would mean showing that "IBF" children do not do better just because their married parents have higher incomes, but that they do better even when the incomes of their households and the households of homosexual parents are the same. Again, Regnerus has done these comparisons for "LMs" (children of "lesbian mothers") and "GFs" (children of gay fathers) separately.
There are eight outcome variables where differences between the children of homosexual parents and married parents were not only present, and favorable to the married parents, but where these findings were statistically significant for both children of lesbian mothers and "gay" fathers and both with and without controls. While all the findings in the study are important, these are the strongest possible ones--virtually irrefutable. Compared with children raised by their married biological parents (IBF), children of homosexual parents (LM and GF):
Are much more likely to have received welfare (IBF 17%; LM 69%; GF 57%)
Have lower educational attainment
Report less safety and security in their family of origin
Report more ongoing "negative impact" from their family of origin
Are more likely to suffer from depression
Have been arrested more often
If they are female, have had more sexual partners--both male and female
The high mathematical standard of "statistical significance" was more difficult to reach for the children of "gay fathers" in this study because there were fewer of them. The following, however, are some additional areas in which the children of lesbian mothers (who represented 71% of all the children with homosexual parents in this study) differed from the IBF children, in ways that were statistically significant in both a direct comparison and with controls. Children of lesbian mothers:
Nearly 15,000 people were "screened" for potential participation in the study; in the end almost 3,000, a representative sample, actually completed the survey questionnaire. Of these, 175 reported that their mother had a same-sex romantic relationship while they were growing up, and 73 said the same about their father. These are numbers just large enough to make some statistically robust conclusions in comparing different family structures.
What the Study Found
The study looked at 40 different outcomes, but reported data for children with "lesbian mothers" and those with "gay fathers" separately. Therefore, there actually were 80 outcome measures that could be said to compare children with "homosexual parents" to those from other family structures. When compared with outcomes for children raised by an "intact biological family" (with a married, biological mother and father), the children of homosexuals did worse (or, in the case of their own sexual orientation, were more likely to deviate from the societal norm) on 77 out of 80 outcome measures.(The only exceptions: children of "gay fathers" were more likely to vote; children of lesbians used alcohol less frequently; and children of "gay fathers" used alcohol at the same rate as those in intact biological families).
Of course, anyone who has had a college course in statistics knows that when a survey shows there are differences between two groups, it is important to test whether that finding is "statistically significant." This is because it is always possible, by chance, that a sample may not accurately reflect the overall population on a particular point. However, through statistical analysis researchers can calculate the likelihood of this, and when they have a high level of confidence that a difference identified in the survey represents an actual difference in the national population, we say that finding is "statistically significant." (This does not mean the other findings are unimportant--just that we cannot have as high a level of confidence in them.)
Regnerus has analyzed his findings, and their statistical significance, in two ways--first by a simple and direct comparison between what is reported by the children of homosexual parents and the children of "intact biological families" ("IBFs"), and second by "controlling" for a variety of other characteristics. "Controlling for income," for example, would mean showing that "IBF" children do not do better just because their married parents have higher incomes, but that they do better even when the incomes of their households and the households of homosexual parents are the same. Again, Regnerus has done these comparisons for "LMs" (children of "lesbian mothers") and "GFs" (children of gay fathers) separately.
There are eight outcome variables where differences between the children of homosexual parents and married parents were not only present, and favorable to the married parents, but where these findings were statistically significant for both children of lesbian mothers and "gay" fathers and both with and without controls. While all the findings in the study are important, these are the strongest possible ones--virtually irrefutable. Compared with children raised by their married biological parents (IBF), children of homosexual parents (LM and GF):
Are much more likely to have received welfare (IBF 17%; LM 69%; GF 57%)
Have lower educational attainment
Report less safety and security in their family of origin
Report more ongoing "negative impact" from their family of origin
Are more likely to suffer from depression
Have been arrested more often
If they are female, have had more sexual partners--both male and female
The high mathematical standard of "statistical significance" was more difficult to reach for the children of "gay fathers" in this study because there were fewer of them. The following, however, are some additional areas in which the children of lesbian mothers (who represented 71% of all the children with homosexual parents in this study) differed from the IBF children, in ways that were statistically significant in both a direct comparison and with controls. Children of lesbian mothers:
Are more likely to be currently cohabiting
Are almost 4 times more likely to be currently on public assistance
Are less likely to be currently employed full-time
Are more than 3 times more likely to be unemployed
Are nearly 4 times more likely to identify as something other than entirely heterosexual
Are 3 times as likely to have had an affair while married or cohabiting
Are an astonishing 10 times more likely to have been "touched sexually by a parent or other adult caregiver."
Are nearly 4 times as likely to have been "physically forced" to have sex against their will
Are more likely to have "attachment" problems related to the ability to depend on others
Use marijuana more frequently
Smoke more frequently
Watch TV for long periods more frequently
Have more often pled guilty to a non-minor offense
Differences in Sexuality
When comparing children of homosexuals with children of married biological parents, the differences in sexuality--experiences of sexual abuse, number of sexual partners, and homosexual feelings and experiences among the children themselves--were among the most striking. While not all of the findings mentioned below have the same level of "statistical significance" as those mentioned above, they remain important.
At one time, defenders of homosexual parents not only argued that their children do fine on psychological and developmental measures, but they also said that children of homosexuals "are no more likely to be gay" than children of heterosexuals. That claim will be impossible to maintain in light of this study. It found that children of homosexual fathers are nearly 3 times as likely, and children of lesbian mothers are nearly 4 times as likely, to identify as something other than entirely heterosexual. Children of lesbian mothers are 75% more likely, and children of homosexual fathers are 3 times more likely, to be currently in a same-sex romantic relationship.
The same holds true with the number of sexual partners. Both males and females who were raised by both lesbian mothers and homosexual fathers have more opposite-sex (heterosexual) partners than children of married biological parents (daughters of homosexual fathers had twice as many). But the differences in homosexual conduct are even greater. The daughters of lesbians have 4 times as many female (that is, same-sex) sexual partners than the daughters of married biological parents, and the daughters of homosexual fathers have 6 times as many. Meanwhile, the sons of both lesbian mothers and homosexual fathers have 7 times as many male (same-sex) sexual partners as sons of married biological parents.
The most shocking and troubling outcomes, however, are those related to sexual abuse. Children raised by a lesbian mother were 10 times more likely to have been "touched sexually by a parent or other adult caregiver" (23% reported this, vs. only 2% for children of married biological parents), while those raised by a homosexual father were 3 times more likely (reported by 6%). In his text, but not in his charts, Regnerus breaks out these figures for only female victims, and the ratios remain similar (3% IBF; 31% LM; 10% GF). As to the question of whether you have "ever been physically forced" to have sex against your will (not necessarily in childhood), affirmative answers came from 8% of children of married biological parents, 31% of children of lesbian mothers (nearly 4 times as many), and 25% of the children of homosexual fathers (3 times as many). Again, when Regnerus breaks these figures out for females (who are more likely to be victims of sexual abuse in general), such abuse was reported by 14% of IBFs, but 3 times as many of the LMs (46%) and GFs (52%).
These data require more detailed exploration and explanation. A number of researchers have pointed out that self-identified homosexual adults (both men and women) are more likely to report having been victims of child sexual abuse. However, Family Research Council and other pro-family organizations have been criticized for also pointing to evidence suggesting that homosexual men are more likely to commit acts of child sexual abuse than are heterosexual men. And experts in child sexual abuse in general say that men are most often the perpetrators, regardless of the sex of the victim. Therefore, the finding that children of lesbian mothers are significantly more likely to have been victims of sexual touching by "a parent or adult caregiver" than even the children of homosexual fathers is counter-intuitive.
However, it is important to note what we do not know about such experiences from the data that have been published. The fact that a child of a lesbian mother was touched by "a parent or adult caregiver" does not mean that the lesbian mother was herself the parent or caregiver who did the "touching." An alternative scenario mentioned by Regnerus, for example--hypothetical, but plausible--is one in which a child is molested by her biological father; her mother divorces her father; and the mother later enters into a lesbian relationship.
Limitations of the Study
While the Regnerus study is a vast improvement over virtually all the prior research in the field, it still leaves much to study and learn about homosexual parents and their effect on children. Author Mark Regnerus emphasizes the traditional caveat in social science, warning against leaping to conclusions regarding "causality." In other words, just because there are statistical correlations between having a homosexual parent and experiencing negative outcomes does not automatically prove that having a homosexual parent is what caused the negative outcomes--other factors could be at work.
This is true in a strict scientific sense--but because Regnerus carefully controlled for so many other factors in the social environment, the study gives a clear indication that it is this parental characteristic which best defines the household environment that produces these troubling outcomes. The large number of significant negative outcomes in this study gives legitimate reason for concern about the consequences of "homosexual parenting."
The definition of what it means to have a homosexual parent is also a loose one in this study--by necessity, in order to maximize the sample size of homosexual parents. Not all of those who reported that a parent was in a same-sex relationship even lived with that parent during the relationship; many who did, did not live with the partner as well. Only 23% of those with a lesbian mother, and only 2% of those with a homosexual father, had spent as long as three years living in a household with the homosexual parent and the parent's partner at the same time. Details like this involving the actual timeline of these children's lives can reportedly be found in Regnerus' dataset, which is to be made available to other researchers later this year.
Figures like these suggest a need for more research, to distinguish, for example, the effects of living with a homosexual parent from having a non-custodial one, or the effects of living with a homosexual single parent vs. a homosexual couple. But they also point out something of note for public policy debates on "gay families"--the stereotype put forward by pro-homosexual activists, of a same-sex couple jointly parenting a child from birth (following either adoption or the use of artificial reproductive technology), represents a scenario that is extraordinarily rare in real life. Most "homosexual parents" have their own biological children who were conceived in the context of a previous heterosexual relationship or marriage, which then ended before the person entered into homosexual relationships.
Conclusion
The articles by Marks and Regnerus have completely changed the playing field for debates about homosexual parents, "gay families," and same-sex "marriage." The myths that children of homosexual parents are "no different" from other children and suffer "no harm" from being raised by homosexual parents have been shattered forever.
[1] Mark Regnerus, "How different are the adult children of parents who have same-sex relationships? Findings from the New Family Structures Study," Social Science Research Vol 41, Issue 4 (July 2012), pp. 752-770; online at: http://www.sciencedirect.com/science/article/pii/S0049089X12000610
[2] Loren Marks, "Same-sex parenting and children's outcomes: A closer examination of the American Psychological Association's brief on lesbian and gay parenting," Social Science Research Vol 41, Issue 4 (July 2012), pp. 735-751; online at: http://www.sciencedirect.com/science/article/pii/S0049089X12000580
