Cornell University Law School Social Science and Law

Same-Sex Marriage: The Expert Evidence of Dr. Mark Regnerus in DeBoer v. Snyder

Compiled by Jialu Li, Matthew Tinker, and Spencer Teplitz

The Expert: Dr. Mark Regnerus

(1) The expert. First, how would you evaluate the background, credentials, and other qualifications of the expert who is the focus of your project?

Mark Regnerus has a Phd. from UNC and is a professor of sociology at the University of Texas. In addition. Dr. Regnerus has authored multiple books and other publications on topics related to sexual decision-making. This case is the first time Dr. Regnerus has served as an expert. Lastly, he has at least a partial background in studies on same-sex couples. 

Were there challenges to the neutrality of the social scientist conducting the research or serving as an expert? Should there have been?

One issue the court found with the neutrality of Dr. Regnerus’s evidence was that it was funded by the Witherspoon Institute and the Bradley Foundation. These two private organizations are well known for their political conservatism. Whether the support from these organizations influenced Dr. Regnerus’s study or not is difficult to measure. However, the court was well aware of who supported the study and factored that in when deeming it uncredible. 

The Social Science Research

Evaluating the scientific merit of Dr. Regnerus’s research involves (1) studying the research structure and methodology at a high level; (2) critically examining the research through the use of social science tools; and (3) surveying responses from relevant science communities for guidance on the quality of the underlying research.

In his expert report filed in Deboer v. Snyder, Regnerus describes the New Family Structures Study (“NFSS”) as a “survey data collection project that screened over 15,000 young adults about their childhood structure, and collected complete surveys from just under 3,000 of them, including 248 responses who reported a parental same-sex romantic relationship.”  Expert Report of Mark D. Regnerus at 3, Deboer v. Snyder, 973 F. Supp. 2d 757 (E.D. Mich. 2014) (No. 12-CV-10285), rev’d, 772 F.3d 388 (6th Cir. 2014), rev’d sub nom. Obergefell v. Hodges, 574 U.S. 1118 (2015).  Regnerus sought to distinguish his study from past studies on the topic of same-sex parenting that had involved “non-random, non-representative samples, often with relatively few participants” and that often used “comparison groups of reduced kinship, such as stepfamilies.”  Id.  

The study published by Regnerus based on the NFSS data is a cross-sectional study that examined behavioral outcomes, as reported by the adult children, across several groups.  Mark Regnerus, How Different Are the Adult Children of Parents Who Have Same-Sex Relationships? Findings from the New Family Structures Study, 41 Soc. Sci. Res. 752, 753 (2012) [hereinafter Findings].   The NFSS instead used a “large, representative pool drawn from a population-based random sample.”  Expert Report of Mark Regnerus, Deboer, 973 F. Supp. 2d 757 (No. 12-CV-10285), at 7.  Regnerus has argued that this method avoids significant issues involved with snowball sampling, such as community bias and non-random pools, that might negatively impact the validity of past studies on this topic.  See Regnerus, Findings at 755.  Because the NFSS was not a longitudinal study, it did not purport to address causation.  Id.  Regnerus testified that the study’s goal was focused more narrowly on testing for the existing of “simple group differences, and--with the addition of several control variables--to assess just how robust any between-group differences were,” without reaching the issue of the source of any particular group difference that the survey might uncover.  Expert Report of Mark Regnerus, Deboer, 973 F. Supp. 2d 757 (No. 12-CV-10285), at 7.

At the direction of Dr. Regnerus’ employer, the University of Texas at Austin, data collection for the NFSS was conducted by Knowledge Networks (“KN”), a research firm commonly used by other organizations working in the field as well as by Plaintiff’s expert Michael Rosenfeld.  See Regnerus, Findings at 755–56; Expert Report of Mark Regnerus, Deboer, 973 F. Supp. 2d 757 (No. 12-CV-10285), at 12.  KN assembled an online research panel that was representative of the US population by recruiting randomly through the use of telephone and mail surveys.  Id. at 756.  

The NFSS collected from the respondents by asking about the marital status of their biological parents.  Combining responses about the same-sex romatic behavior of parents and levels of household stability, Regnerus structured this data according to eight types of family-of-origin structure:

1. IBF: Lived in intact biological family (with mother and father) from 0 to 18, and parents are still married at present …
  
2. LM: R reported R’s mother had a same-sex romantic (lesbian) relationship with a woman, regardless of any other household transitions …
  
3. GF: R reported R’s father had a same-sex romantic (gay) relationship with a man, regardless of any other household transitions …
  
4. Adopted: R was adopted by one or two strangers at birth or before age 2 …
 
5. Divorced later or had joint custody: R reported living with biological mother and father from birth to age 18, but parents are not married at present …
 
6. Stepfamily: Biological parents were either never married or else divorced, and R’s primary custodial parent was married to someone else before R turned 18 …
 
7. Single parent: Biological parents were either never married or else divorced, and R’s primary custodial parent did not marry (or remarry) before R turned 18 …
 
8. All others: Includes all other family structure/event combinations, such as respondents with a deceased parent …

Id. at 756.  In his comparative assessments on outcomes, Regnerus used the IBF group, that is, stable heterosexual couples, as the control group.  See Expert Report of Mark Regnerus, Deboer, 973 F. Supp. 2d 757 (No. 12-CV-10285), at 31.  While separating out a variety of non-stable heterosexual parental relationships in categories 4-8, Regnerus collapsed all same-sex parental relationships into single categories 2-3.  See id.
  
Across a range of dichotomous outcome variables (such as “currently employed full-time,” and “voted in last presidential election”), continuous outcome variables (such as “educational attainment,” and “self-reported overall happiness”), and event-count outcome variables (such as “frequency of marimuana use,” and “frequency of having been arrested”), Regnerus reported optimal outcomes, where that could be facially determined, to be found in the IBF group.  See Regnerus, Findings at 761–62.  The comparisons found statistically significant differences both between IBFs and LMs and between IBFs and GFs in directions that were clearly “suboptimal”.  See id. at 764–765.  Regnerus noted these findings diverged from past studies and assigned the difference partially to the small or nonprobability samples used in previous studies.  See id. at 765.  He argued that his study avoided sample-size bias amd accounted for a broader diversity of same-sex household experiences.  Id.

We argue that Regnerus’s testimony would not have been admitted using the Frye “general acceptance” test.  Plaintiff notes that in the same issue of Social Science Research in which the Regenerus study was published, the issue also included a letter signed by over one hundred social scientists criticizing the Regenerus study for “failing to follow standard peer-review requirements and for failing to take account of family structure and family instability.”  Pl.’s Mot. In Lim. to Exclude Test. of Mark Regnerus, Deboer, 973 F. Supp. 2d 757 (No. 12-CV-10285), at 18–19.  The American Sociological Association (“ASA”) found a number of significant errors in the Regnerus report, most notably:

First, Regnerus … does not examine children born or adopted into same-sex parent families, but instead examines children who, from the time they were born until they were 18 or moved out, had a parent who at some time had “a same-sex romantic relationship.” … Regnerus noted that “just under half” of the individuals characterized by him as children of ‘lesbian mothers’ and ‘gay fathers’ were the offspring of failed differentsex marriages whose parent subsequently had a samesex relationship. … In other words, the study did not analyze children of two same-sex parents.

Second, Regnerus … compared (i) the children of parents who at some point had a “same-sex romantic relationship,” and who had experienced a family dissolution or single motherhood, to (ii) children raised by two biological, married different-sex parents. The study also stripped away all divorced, single, and stepparent families from the different-sex parent group, leaving only stable, married, different-sex parent families as the comparison. … It was hardly surprising that this select different-sex parent group had better outcomes given that stability … is a key predictor of child wellbeing. By removing divorced, single, and step-parent families from the different-sex parent group, the Regnerus … paper makes inappropriate apples-to-oranges comparisons.

Brief for American Sociological Association as Amici Curiae Supporting Plaintiffs, Deboer, 973 F. Supp. 2d 757 (No. 12-CV-10285), at 21–23.  The ASA also noted the study did not consider the duration of time of a parent’s same-sex romantic involvement and did not isolate whehther the recorded events occurred during the time the respondent lived with the parents’ same-sex partner or during a different time.  See id.

Perhaps, most significantly, the journal that first published the study, in a subsequent internal audit, concluded that Regenerus’s study was non-scientific, had “serious flaws and distortions,” and should not have been published.  See Darren E. Sherkat, The Editorial Process and Politicized Scholarships: Monday Morning Editorial Quarterbacking and a Call for Scientific Vigilance, 41 Soc. Sci. Res. 1346, 1346–49.  In its response, the State did not provide any documentation of general acceptance of the study's methodology and simply asserted in the alternative that the study had been generally accepted.  See State Def. Resp. in Opp’n to Pl.’s Mot. In Lim. to Exclude Test. of Mark Regnerus, Deboer, 973 F. Supp. 2d 757 (No. 12-CV-10285), at 8.  Due to the lack of evidence presented to the contrary, we conclude that the expert testimony of Mark Regnerus would not pass Frye gatekeeping.

Dabuert provided four criteria to use in measuring the validity of the underlying social science research: (1) whether the expert's technique or theory can be or has been tested---that is, whether the expert's theory can be challenged in some objective sense, or whether it is instead simply a subjective, conclusory approach that cannot reasonably be assessed for reliability; (2) whether the technique or theory has been subject to peer review and publication; (3) the known or potential rate of error of the technique or theory when applied; (4) whether the technique or theory has been generally accepted in the scientific community.  See Daubert v. Merrell Dow Pharm., 509 U.S. 579, 593–96.  We find points 1, 2, and 4 to be most relevant for this case.  On the first point of whether Regnerus’s opinion has been tested, as described in the above ASA overview, the methodology been reviewed and found lacking in a number of key ways.  On the second point, although the study was published, the publishing journal has since stated that publishing the study had been an error.  Finally, as noted in the above analysis of Frye gatekeeping, Regnerus’s theory and techniques have not found general acceptance.  Therefore, we believe the test should fail Daubert gatekeeping.

Although the court found to the contrary in this case, we believe the the probative value of the expert testimony could have been excluded in this case, under Frye or Daubert gatekeeping, because its probative value is substantially outweighed by unfair prejudice, confusing of the issues, and the needless presentation of cumulative evidence.  See Fed. R. Evid. 403.  Although the State argued that any limitations of the study should go to weight instead of to inadmissibility, we believe that the fact that study was prepared in anticipation of the litigation should have tilted the court into simply not allowing the study in the first place.

Relevance to the legal issues in the case.

Part 3. Relevance to the legal issues in the case.
 
Was the expert evidence relevant to the central legal questions in the case? 

The central issue in the Deboer case is ultimately whether same-sex couples are should be banned from getting married under a rational basis standard of review. One sub issue central to Deboer is the whether same-sex couples could jointly adopt children.  Mark Regenrus’s expert testimony, as poorly as it was conducted, was relevant to the sub issue of same-sex couples jointly addopting children. Regenrus’s studies attempted to show that same-sex couples raised children who grew up to be misfits and troublemakers in society. The topic of Regenrus’s study is relevant to the legal question but is done so poorly that it leaves the court to question the validity of the evidence.
 
What is your conclusion about the importance of the social science evidence to the major claims being adjudicated?

The social science evidence could have been much more important to the main legal question if it was done properly. The court points out many red flags in Regnerus’s testimony. For example the court noted, “Regnerus reported that "just over half (90) of the 175 respondents whose mother had a lesbian relationship reported that they did not live with both their mother and her same-sex partner at the same time." Id. at 11. In addition, the study was funded by a biased source who may have driven the study a certain direction. Id. If the study was not so misleading perhaps it could have been somewhat useful during the adjudication of the legal issues. 

The Court’s Treatment of Dr. Regnerus’s Evidence

The court discussed the methodology of the expert evidence and laid out criticism against Regnerus’s testimony. First, the research “failed to measure the adult outcomes of children who were actually raised in same-sax households.” Deboer, 973 F. Supp. 2d 757 (No. 12-CV-10285), at 21. This is a big problem of the research because among respondents many parental same-sex romantic relationships lasted for only brief periods of time and many of the participants never lived in a same-sex household at all. Only respondents with experience of parental same-sex romantic relationships could not state potential outcomes of children-raising in same-sex marriage families. Second, critics have concerns that “the NFSS made an unfair comparison between children raised by parents who happened to engage in some form of same-sex relationship and those raised by intact biological families.” Id.

The court also stated that “even Regnerus recognized the limitations of the NFSS.” Id. Regnerus also testified that “there is no conclusive evidence that ‘growing up in households wherein parents are in (or have been in) same-sex relationships’ does not adversely affect child outcomes.” Id., at 22.

What is more, the court found that Regnerus’s 2012 study was hastily concocted at the behest of a third-party funder whose concept of the institution of marriage was threatened by court decisions. While Regnerus maintained that the funder did not affect his impartiality in the research, the court found the testimony unbelievable. Instead of comparing outcomes of children raised by same-sex couples with those of children raised by heterosexual couples, Regnerus equated being raised by a same-sex couple with having ever lived with a parent who had a "romantic relationship with someone of the same sex" for any length of time. Id., at 24. The court concluded that Regnerus’s testimony “entirely unbelievable and not worthy of serious consideration.” Id.

The judge accurately interpreted the study and rejected the evidence appropriately.

The circuit court agreed with district court with regard to the evidence that all state’s studies were “given little credence by the district court because of inherent flaws in the methods used or the intent of authors.” Id., at 22.
The appellate court discussed the NFSS research by Regnerus as an example and stated that Regnerus conceded that “his own department took the highly unusual step of issuing the following statement on the university website in response to the release of the study:

            [Dr. Regnerus's opinions] do not reflect the views of the sociology department of the University of Texas at Austin. Nor do they reflect the views of the American Sociological Association which takes the position that the conclusions he draws from his study of gay parenting are fundamentally flawed on conceptual and methodological grounds and that the findings from Dr. Regnerus'[s] work have been cited inappropriately in efforts to diminish the civil rights and legitimacy of LBGTQ partners and their families.” Deboer, 772 F.3d 388., at 426.

The circuit court also pointed out that the district court reflected clearly that Regnerus’s study had been funded by a conservative institute opposed to same-sex marriage.