Saving Faith and Expertise

This guest post is by Kevin Shafer, an Associate Professor of Sociology at Brigham Young University and an adjunct associate professor of Health & Society at McMaster University (Canada). He holds a PhD in Sociology from The Ohio State University. His scholarship focuses on mental health and father involvement in a cross-national perspective.

In October 2018, President Dallin H. Oaks cautioned that “expertise in one field should not be taken as expertise on truth in other subjects.” In his new book, Saving Faith, Egyptologist John Gee makes assertions about child abuse victimization, LGBTQ+ identity, and the potential for child abuse victims to become perpetrators of abuse in adulthood. These are questions that are central to social science and strong claims are being made by someone without training in psychology, sociology, social work, economics, or related disciplines. Professor Gee’s lack of expertise in these areas is may have led him to make errors that lead to problematic claims that are not born out by research. In contrast to Professor Gee, I am a sociologist that researches gender, mental health (including childhood adversity), and family life. Here, I discuss two claims made in the book and why they are not based in science or current church statements on sexuality.

Childhood Abuse & Sexuality
On page 195 of his book, Professor Gee writes the following on the positive correlation between LGBTQ+ identity and childhood sexual abuse: “[n]onheterosexuals are about twice as likely as heterosexuals to have suffered emotional, physical, or sexual abuse as children. While ‘the vast majority of individuals who suffer childhood trauma do not become gay or bisexual’ it is still a viable hypothesis ‘that adverse childhood experiences may be a significant—but not determinative—factor in developing homosexual preferences.’” Gee quotes an outside source and we should consider the foundation for his claim. Gee’s quote comes from a literature review by Mayer and McHugh in the journal The New Atlantis[1]an outlet that does not adhere to any agreed-upon standard of rigour in the social sciences. The New Atlantis is not peer-reviewed and was affiliated with the Ethics and Public Policy Center (EPPC) when the Mayer and McHugh report was published. EPPC indicates on their webpage that they are dedicated to “applying Judeo-Christian moral tradition to critical issues of public policy.” Mayer and McHugh have been roundly criticized by many, including National Institutes of Health geneticist Dean Hammer, who called the report a “selective and outdated collection of references and arguments aimed at confusing, rather than clarifying our understanding of sexual orientation and gender identity.”

Gee’s quote from Mayer and McHugh’s is based on a 2011 meta-analysis by Friedman and colleagues in the American Journal of Public Health[2], a highly respected scholarly outlet. For the unfamiliar, meta-analyses use statistical methods to generate summaries from many scholarly studies. In this case, Friedman et al. summarize results from 49 studies on sexual identity and childhood sexual abuse conducted in the United States during the 1990s and 2000s. Indeed, they find that LGBTQ+ individuals were 3.8 times more likely to experience childhood sexual abuse, 1.2 times more likely to experience parental physical abuse, 1.7 times more likely to experience assault at school, and 2.4 times more likely to skip school out of fear. Yet, these authors come to very different conclusions than Gee or Mayer and McHugh, writing that: “sexual abuse does not cause individuals to become gay, lesbian, or bisexual. Sexual minority individuals are instead more likely to be targeted for sexual abuse, as youths who are perceived to be gay, lesbian, or bisexual are more likely to be bullied by their peers.” Another potential issue, not raised by the authors, is reporting bias. Heterosexual respondents, particularly males, may be less likely to report sexual abuse due to strong masculine norms around sexuality. Indeed, Boston College social work professor Scott Easton found that the median time to abuse disclosure was more than 20 years in men, but less than one year for women.

Notably, Northwestern psychologist Michael Bailey and colleagues[3] wrote in a 2016 meta-analysis that although “no specific theory of what causes people to be attracted to men, to women, or to both has received enough support to win the backing of all reasonable scientists” but noted that the current preponderance of evidence suggests that gay, lesbian, or bisexual identity is much more likely due to a “so-far-unspecified reason of nature rather than social nurture.” In contrast, Gee relies on a review by an unreputable source that is willfully misinterpreting scholarly findings. Although Gee argues that abuse is not “determinative” he does suggest it may be “significant.” Yet, even in a generous reading, it seems clear that Gee has not done the necessary homework on a controversial claim that requires a scholar to do a careful, critical investigation of the research literature. Failure to do so perpetuates harmful ideas about sexual minorities, particularly when scholars agree that LGBTQ+ are victimized because of their sexual identities.

Experiencing & Perpetuating Sexual Abuse
Gee’s second claim of concern also appears on page 197 where he writes that “victims of sexual abuse deserve compassion and assistance. They also deserve special care because they are more likely to become sexual abusers of children.” Here, Gee cites a single study of 490 Germans[4]. To be fair, the cited study is written in German—a language that I do not read, write, or speak. Nevertheless, I am very familiar with this research literature, given that a significant strain of my research focuses on Adverse Childhood Experiences (ACEs), including childhood sexual abuse (CSA), and parenting behavior.  Gee’s argument that victims are more likely to become abusers is a controversial one. Indeed, his claim is one built on methodologically problematic research. Many studies use retrospective reports that may unreliably measure sexual abuse victimization—particularly among those that indicate no history. Others use data from child welfare agencies that systemically target low income and minority individuals who are a much higher risk of experiencing abuse and neglect as a child. There may be significant selection bias in these studies. Concern about individuals that have disclosed sexual abuse histories may also be the result of detection bias. Victims of CSA may be monitored more strongly because of widespread ideas that “victims of sexual abuse…are more likely to become sexual abusers.” Data that follows respondents from childhood into adulthood, however, shows no statistically significant relationship between CSA and being abusive[5]. My own work, with Scott Easton of Boston College, finds no relationship between victimization and abusing in American fathers[6]. Strong causal claims that fail to acknowledge or disclose mixed findings or methodological issues have the potential to perpetuate harmful, marginalizing stereotypes about victims of sexual abuse.

This is not to minimize the evils of child abuse. It is deeply troubling, causing great harm to individuals and communities. CSA victims are at increased risk for depression, anxiety, post-traumatic stress disorder, poor physical health, poor economic outcomes, and often lack of loving, supportive social connections. The negative effects are broad and, in this way, Gee is correct: “victims of sexual abuse deserve compassion and assistance.” Unfortunately, claims that lack appropriate nuance have the potential to undermine such efforts. Indeed, CSA victims are extremely unlikely to become abusers themselves—particularly if they get quick, responsive help at an early age. We should do more to destigmatize victimization so that more are willing to speak up. This stigma is strong. I’m afraid that claims like those made by Gee will perpetuate the stigmatization of victimization and prevent many, particularly men, from seeking out the help that they desperately need.

Leave Social Science to the Social Scientists
When researchers fail to properly report nuance, make strong claims based on limited reading, and fail to acknowledge the shortcomings of social research, they can cause real harm. In this case, the marginalization of LGBTQ+ individuals can perpetuate the very abuses that Gee shows concern for, given that anti-LGBTQ+ bias appears to be a significant correlate of their victimization. Stigmatizing abuse victims may generate significant psychological, physical, social, and developmental harm. Perhaps we should take President Oaks’ advice to heart—both as scholars and readers.

[1] Mayer, L. S., & McHugh, P. R. (2016). Sexuality and gender: Findings from the biological, psychological, and social sciences. The New Atlantis, 10-143.

[2] Friedman, M. S., Marshal, M. P., Guadamuz, T. E., Wei, C., Wong, C. F., Saewyc, E. M., & Stall, R. (2011). A meta-analysis of disparities in childhood sexual abuse, parental physical abuse, and peer victimization among sexual minority and sexual nonminority individuals. American Journal of Public Health101(8), 1481-1494.

[3] Bailey, J. M., Vasey, P. L., Diamond, L. M., Breedlove, S. M., Vilain, E., & Epprecht, M. (2016). Sexual orientation, controversy, and science. Psychological Science in the Public Interest17(2), 45-101.

[4] Urban, D., & Fiebig, J. (2011). Pädosexueller Missbrauch: wenn Opfer zu Tätern werden/Pedosexual Abuse: When Victims Become Abusers: Eine empirische Studie/An Empirical Study. Zeitschrift für Soziologie40(1), 42-61.

[5] Widom, C. S., Czaja, S. J., & DuMont, K. A. (2015). Intergenerational Transmission of Child abuse and Neglect: Real or Detection Bias? Science, 347(6229), 1480–1485.

[6] Shafer, K. & Easton, S.D. (Under review). “Adverse Childhood Experiences, Psychological Distress, & Fathering Behaviors.” Journal of Marriage & Family.


  1. I’ve reread your paragraph twice, but I don’t see “willfully misinterpreted” spelled out clearly nor evidence presented for such willfulness; I do not see Gee making monocausal claims of “cause” i.e. “sexual abuse cause[s] individuals to become gay, lesbian, or bisexual.” Nor does Gee does not appear to be expressing “a conclusion.” At least as you present it, that looks as much like you overreading Gee as much as you present Gee as overreading theses studies.

    “Gee’s argument that victims are more likely to become abusers is a controversial one.” By controversial, you intend “heavily contested because the data are not yet clear,” yes?

    Thanks for clarifying

  2. Rather, “Nor does Gee appear to be expressing a conclusion.”

  3. Gee’s second claim of concern also appears on page 197 where he writes that “victims of sexual abuse deserve compassion and assistance. They also deserve special care because they are more likely to become sexual abusers of children.” Here, Gee cites a single study of 490 Germans.

    If the quotation is accurate, it overstates the case made in the cited article.

    The study in question looks at 130 convicted pedophiles who at the time of the study were in prison. The rest belong to two control groups (offenders who committed sexual crimes against people older than 14 and offenders who committed non–sexual violent crimes). And while the abstract mentions 490 offenders, the study (page 46) itself refers to a total of just 354 individuals belonging to the three groups. Whether 490 or 354 total, the group of convicted pedophiles is a smaller subset.

    At any rate, the authors of the study go to great lengths to limit the applicability of their findings that, indeed, sexual abuse of a child increases the risk of becoming a pedophile. First, the analysis does not apply to active pedophiles generally, just the specific population (the 130 convicted and imprisoned pedophiles) in the study. Second, the relevance of the results of the study to any other population cannot be determined empirically. Third, the results are subject to methodological limitations—they depend on assumptions that cannot always be empirically verified. Fourth, the results are the product of a moment in time and can change over time. Finally, the authors acknowledge that the measurement results could be biased. All this can be found on page 57.

    These caveats by the authors—even without critiquing the study further—should be enough to give anyone pause with regard to Gee‘s conclusion in his book quoted above.

  4. I did not accuse Gee of willfully misinterpreting. I accused the article he cites of willfully misinterpreting an article because it writes it interprets findings in the exact opposite way of a paper it uses to make its claims.

    And yes, controversial because the data are far from conclusive. And a lack of nuanced discussion on these issues is problematic.

  5. Gee has no place in acting like an expert in sociology.
    He is not even a honest expert in his own field of Egyptology.
    He has already discredited his own name.
    It is a money grab to get orthodox Deseret Book shoppers to spend.

  6. And now it looks like the book has been pulled from Deseret Book: (404 Error)

  7. I take it you read the entire chapter? And the section right before the infamous page number that is being passed around on social media; a section that is dedicated specifically to causes of homosexuality that relies heavily on the Bailey et al. study you quote? Because if you didn’t, then you are indeed overreading Gee. Granted, he may have needed to be more nuanced on this particular point instead of relying solely on the quote from McHugh and Mayer. And if this was the only thing he said about causes of homosexuality in the book, then the outrage would be understandable. But he literally just spent four pages providing different potential explanations. He concludes,

    “To sum up, so far studies have shown neither one specific cause of homosexuality nor a specific factor that precipitates homosexuality in most cases when it is present. It seems likely that there are a number of different ways that an individual can come to consider themselves homosexual” (pg. 194).

    Readers aren’t suddenly going to have amnesia once they hit the abuse section.

    As for claims about Gee’s dishonest expertise in Egyptology, his CV would beg to differ.

  8. How embarrassing! Who agreed to publish his work with such heinous and harmful errors? Thanks for correcting and for sharing insight into how lgbt kids can be targets for abuse.

  9. If he fully read Bailey et al., then he would know that they also note that his argument about sexual abuse and being LGBTQ+ is in the wrong direction. Since they dedicate more than a journal article page to the issue, agreeing with the conclusions of Friedman, et al.

  10. This is a tacit admission that you didn’t read the chapter. You took a page off social media and decided to blog about it. Either that or someone else who also hasn’t read the chapter contacted you to come refute Gee. Context matters.

  11. “Because a number of different causes and effects are combined in this study, specific causation is not demonstrable and so the results need to be interpreted cautiously” (pg. 195).

    This is the sentence right before the one you quote from Gee. Funny how that got left out.

  12. You mean the chapter where he indicates people’s sexuality when referencing their research, talks at length about problematic studies, then provides a single sentence indicating that some other research doesn’t find that result? That chapter?

    I discuss these issues because I study child abuse and mental health. The claims are incomplete and the research is shoddy.

  13. Funny how you leave out that the studies have the association going the other direction. Which Gee does not.

  14. He’s also citing something that’s not a study, but a literature review in a non-peer reviewed report—so I don’t even know what that means, given there is no statistical analysis in the review.

  15. In other words, the caveat means nothing if the study is being interpreted in the exact opposite way of that paper. That is the issue. And the not representing a full and complete picture of the literature.

  16. No thanks. He cited research. He didn’t claim to be an expert. You can disagree and cite research done on your own. You don’t need to attack his credible to make a conclusion. That’s stretching Pres Oaks words beyond the mark.

    The fact is there is a strong correlation for many, many sexual deviations and abuse linked with past childhood experiences. The foundation he was standing on is not as shakey as you suggest.

  17. Isn’t writing a book an implicit, if not explicit, claim of expertise in a subject?

  18. Because the study he’s citing is_checks notes_Bailey et al. *Then* he goes on to quote McHugh & Mayer about abuse being a viable hypothesis.

    It’s basically “causation is tricky in these studies, so be careful with the interpretations. But it’s possible that abuse may play a role in some people’s homosexuality.”

  19. Except the very study being cited says it’s *not* a viable hypothesis. This is exactly what I mean by shoddy research. He didn’t go directly to the source, he relied on a report in a non-peer reviewed, non-scientific journal.

  20. He also does not cite Bailey et al, for these particular claims and even if he did, Bailey et al also argue that it’s not a viable hypothesis—so I’m confused what the issue is.

  21. Thank you for your analysis. Professor Gee does the scientific community a disservice by presenting his views in a way that sidesteps the process that gives us confidence in published research. If only President Oaks would also concede that he is not a sociologist and stop flogging his personal hobby horse.

  22. Sorry if this is a threadjack, but what led Friedman et al to conclude that “sexual abuse does not cause individuals to become gay, lesbian, or bisexual. Sexual minority individuals are instead more likely to be targeted for sexual abuse, as youths who are perceived to be gay, lesbian, or bisexual are more likely to be bullied by their peers”?

  23. TT, would you care to weigh in on the conclusion Gee draws from Urban and Fiebig’s findings quoted above? Does Gee qualify that sentence on page 195 elsewhere in the book? I’m happy to extend Gee the benefit of doubt, having only seen page 195, but the way that sentence overstates the authors’s findings in the study he cites, as I pointed out in my comment above, is a red flag that would make me cautious about how well other sources are represented in the book. And while it’s certainly fair game to critique the critic, let’s not lose sight of who published a blog post and who published a book.

  24. Just a note that the person posting as “TT” is not the TT that many long-time bloggers may know by that name. Whether coincidence or impersonation, I don’t know or care, and TT as me has not blogged in many years, but I did want to clear up any confusion.

  25. Thanks for the clarification. Speaking for myself, I had wondered about that.

  26. stephenchardy says:

    These associations between abuse and sexual identity/sexual orientation can be misused, mis-remembered and can result in real pain and grief.

    Most parents of children who come out as LGBTQ+ are shocked to hear it. Most, not all. Even when parents respond generously and with love the news is almost always unwelcome, at least at first. Most parents of LGBTQ+ have never really thought a lot about it before. Oh, of course they have at least thought about it. No one could have made through the last several decades with all of the debate about marriage and other issues without thinking at least a bit about it.

    But for most parents, it was previously an abstract possibility. Once their children’s sexual identity becomes reality most parents go through a careful reckoning about their children, and themselves. On top of all of the difficulties and realigning, they will eventually find the opinions that LGBTQ+ is born out of bad parenting, distant unloving fathers, and sexual, emotional, or physical abuse. Thus, on top of all of the tricky processes that lead to love and acceptance they have to grapple with the idea that their community, their ward, their neighbors, and their family members may think that “they” did it. They abused or hurt their children and “made” them gay. I know of a number of parents of LGBTQ+ who feel alone and isolated at church, who feel the judgment of others, who feel that they are being seen as poor parents.

    And then those who are LGBTQ+ must face not only the prospect of blame being placed on their parents, but must now bear constant scrutiny about their alleged increased potential to be abusers of others.

    So much pain. All so un-necessary. At a time of supreme challenge, can’t we simply offer love and help? Must we lay a pathway for judgment and pain?

  27. Just chiming in to say hey to the OG TT

  28. Thank you, Kevin, for your important corrective.

  29. Thank you, Professor Shafer.

  30. Wait, you really did read only one page before weighing in on someone else’s book, and now you’re on your high horse about what constitutes true scholarly expertise? And you didn’t even read one of the sources he cites because your grad program dropped its language requirement in 1987, or something like that? Dude, that’s a grad student move right there. You’re an assistant professor now, so grow up and be an adult. Sure, you have relevant expertise and you’re well positioned to speak to the specific issue. But you’re massively overreaching to make this about Gee and the misuse of scholarly expertise.

    And your conclusion – leave sociology to the sociologists – is absolutely wrong. The whole point of academic research is to discover things that are useful to people outside the discipline. Sociology gets used in advertising and policy and literary studies and a bunch of other places. This does mean it often gets used badly – such is the nature of interdisciplinary work in any direction – but then academic research also needs to do its part to communicate its findings to make them usable by outsiders. If sociology is relevant to popular works of religious devotion, then you should want authors to consult sociology, not try to claim a monopoly on it for expert sociologists.

  31. The OP being everything I know about John Gee’s work (but not everything I know about the larger topic), I’ll take this line as an accurate portrayal:
    “[I]t is still a viable hypothesis ‘that adverse childhood experiences may be a significant—but not determinative—factor in developing homosexual preferences.’”
    What this says to me is “if you’re in the business of finding non-essential not-inherent or biological reasons for homosexuality, the adverse childhood experience hypothesis has not been rejected.”
    Stated that way it may be technically a true statement. However, I question the use of “viable” for such hypothesis. “Viable hypothesis” suggests a scientific method, whereas this whole approach appears polemical. And if we were operating in the realm of science, or good science at least, I would ask hard questions about the cost of a false positive before glibly attaching “viable.”

  32. But you’re massively overreaching to make this about Gee and the misuse of scholarly expertise.

    No he’s not. I read the study cited by Gee in his, ahem, published book, and it doesn’t do the work Gee says it does. Professor Shafer’s reservations about the conclusion Gee draws from the study are basically shared by the authors of the study he (Shafer) was unable to read.

  33. Ryan Mullen says:

    Thank you, Professor Schafer. These allegations are an embarrassment to BYU’s Religious Studies Center that published Gee’s book. Your response is timely and brings the necessary expertise to bear on evaluating Gee’s claims.

  34. stephencranney says:

    The literature on this subject is a little more nuanced than some might think based on what is being portrayed here. In 2013 Roberts, Glymour, and Koenan published a fairly large (N=34,653) Two-stage least squares study testing the maltreatment-> homosexuality chain of causality, and found evidence for a bidirectional relationship. I haven’t read it in a while, but when I did I remember thinking that their methods and choice of instrumental variables made sense. There’s since been a lot of back-and-forth in Archives of Sexual Behavior on this question since this article (pasted below), and people can definitely disagree, but such disagreements should be rooted in technical aspects of instrumental variable analysis. (Neither here nor there, but one of the explanations Bailey thinks fits the data better is that homosexuality is genetically connected to neuroticism and depression [2014 paper below], which seems just as likely to be damagingly misappropriated by homonegative people as Roberts’ explanation).

    Less technically, all of this isn’t to say that gay people are gay because they were abused, but rather that there is some evidence that at least some of the relationship between abuse and later life orientation runs from abuse to sexual orientation, even if the total amount of homosexuality that is explained by such an effect is minimal compared to the larger “nature” aspects. Framing this as a fringe belief with no evidence that only rightwingers trying to hurt LGBT people hold is not accurate. Sure, Gee shouldn’t have used a source that clearly has an agenda, but there are other sources that he could have used.

    It’s clear, and a doubt Gee or the most conservative of General Authorities would contest this, that nature plays a strong role, but it’s also problematic to assume that it’s 100% nature, and nobody really makes that argument (and I’m not saying that you are, but I thought I’d put that out there). For one the fact that not all people with gay identical twins are gay mitigates against that idea. It’s clear there’s some interaction between nature and nurture that we only fuzzily understand.

    I agree with your take on the second quote.


    Roberts, Andrea L., M. Maria Glymour, and Karestan C. Koenen. “Does maltreatment in childhood affect sexual orientation in adulthood?.” Archives of Sexual Behavior 42, no. 2 (2013): 161-171.

    Original paper

    Bailey, Drew H., and J. Michael Bailey. “Poor instruments lead to poor inferences: Comment on Roberts, Glymour, and Koenen (2013).” Archives of sexual behavior 42, no. 8 (2013): 1649-1652.

    Response to original paper

    Roberts, Andrea L., M. Maria Glymour, and Karestan C. Koenen. “Considering alternative explanations for the associations among childhood adversity, childhood abuse, and adult sexual orientation: Reply to Bailey and Bailey (2013) and Rind (2013).” Archives of sexual behavior 43, no. 1 (2014): 191-196.

    Response to the response to the original paper

    Bailey, Drew H., Jarrod M. Ellingson, and J. Michael Bailey. “Genetic confounds in the study of sexual orientation: Comment on Roberts, Glymour, and Koenen (2014).” Archives of sexual behavior 43, no. 8 (2014): 1675-1677.

    Response to the response to the response to the original paper

  35. peterllc-
    Have you considered there might be a reason that the authors of the study go out of their way to smooth over concerning revelations of their study. Claiming the population of their study can’t be extrapolated, is certainly reasonable statistically — but let’s call a spade a spade; that caveat can be said of virtually every study that isn’t a random sample, double blind, control group, etc.

    Group think is very strong. The authors of the study clearly recognized the conclusions of their study were controversial. Indeed, if they didn’t bend over backwards, pointing to stats101 reasons why their study can’t for certain be applied elsewhere (again those stats101 caveats are true of most studies, right down to the covid ones being justified and applied left and right) their study might not have been published at all.

    Group think is one possibility. Another is, maybe that high correlation is not transferable outside the population being studies (in which case, any credible study always says, “more research is needed in this area” — if more research isn’t called for, then it’s because people don’t want to know the answer). Another is that the authors are telling only what they can tell, and they are allowing useful people like yourself to cover their tracks and people can read between their lines. When someone is being held hostage, a good strategy is to tell the truth through subterfuge.

  36. Remember: sample sizes for mask wearing is in the tens of thousands. For these studies in the hundreds. If you know statistics you’ll understand

  37. Thank you for writing this response to Professor Gee’s book. As a BYU alum and the father of a gay son, I appreciate seeing Professor Gee’s carelessness being corrected and am grateful for your clarity and expertise.

  38. “ Leave Social Science to the Social Scientists”
    This is Orwell’s 1984. Everyone should be encouraged, not discouraged, for doing research.

    Helaman 16:21

    21 And they will, by the cunning and the mysterious arts of the evil one, work some great mystery which we cannot understand, *which will keep us down to be servants to their words, and also servants unto them, for we depend upon them to teach us the word; and thus will they keep us in ignorance if we will yield ourselves unto them* all the days of our lives.

%d bloggers like this: