One of the main arguments against sexual relations between children and adults is the psychological harm caused by sexual abuse. The common opinion is that immense harm is proven by a lot of scientific research.
This is false. There is no research which proves harm related with volitional sex. Only if we mix rape and volitional sex, we obtain statistically significant harm. (Of course, nobody questions that rape and sexual violence are harmful experiences.) Second, the effect size is much smaller than expected in the public opinion.
A variant of this argument claims that sex with children is, if not always harmful, at least risky. This argument is not better - a mixture of volitional sex and rape has of course a risk of being harmful. This does not tell anything about volitional sex.
But, noetheless, to consider the scientific evidence about this mixture allows to extract some interesting evidence.
Richter-Appelt 1994 compares the impact of sexual and physical abuse. Sexual abuse was defined as an unwanted sex involving physical contact, and for physical abuse it was sufficient to mention something like "slaps in the face" regularly. The resulting negative impact was comparable, in general even worse for physical abuse.
Charbonneau & Oxman-Martinez 1996 conclude that "sexual abuse and negligence do not seem to differ fundamentally when it is a matter of establishing the state of health of victims".
Kumar et.al. 1996 have not found any difference between sexually and nonsexually abused adolescents.
In a study of the association of five types of maltreatment (physical abuse, sexual abuse, psychological abuse, neglect, and exposure to family violence)) with a "battery of measures assessing self- and caretaker-reported externalizing and internalizing symptomatology" McGee et.al. 1997 have found that "Psychological maltreatment was the most predictively potent maltreatment type, and enhanced the predictive utility of other maltreatment types." They note that "the findings were consistent with recent research in child maltreatment".
Mullen et.al. 1996, studying the long-term impact of the physical, emotional, and sexual abuse of children, found that "the similarities between the three forms of abuse in terms of their association with negative adult outcomes was more apparent than any differences, though there was a trend for sexual abuse to be particularly associated to sexual problems, emotional abuse to low self-esteem, and physical abuse to marital breakdown. Abuse of all types was more frequent in those from disturbed and disrupted family backgrounds."
Because of the close association between various types of abuse (sexual, physical, emotional) and neglect (physical, emotional) research that does not control these other variables does not prove anything. Often the correlation between sexual abuse and harm disappears if these other variables are controlled. This does not mean that sexual abuse is not harmful, but it seems that large samples are necessary to obtain a statistical significant result: Molnar et al 2001 have found such statistically significant associations between sexual abuse and various types of psychopathology in a survey of 5877 subjects.
See also Salmon & Calderbank 1996; Silverman et.al. 1996
Coffey et.al. 1996 support the thesis that feelings of stigma and self-blame mediate the long-term effects of child abuse.
Literature I have not found interesting: