Cornyn, Paxton Clash Over Controversial Child Case

There was a 10-year-old boy in McLennan County, Texas, who was sexually abused over a period of years by a Waco lawyer named Adam Hoffman (he’s now 14). He testified once. He was cross-examined. He sat in a courtroom while a defense attorney walked him through every inconsistency a frightened child produces under pressure, and then he watched five of 12 jurors conclude that his abuser should walk free. The trial ended in a hung jury. His family then made a decision that no parent should ever have to make, and they made it the way the empirical literature, the prosecutorial-ethics standards, and the common moral sense of any decent person would counsel them to make it. They said no second trial. They said no second cross-examination. They said let this end. Senator John Cornyn now says they were wrong, and he is saying it on television, in press releases, and through political surrogates, in the middle of a U.S. Senate runoff against the Texas attorney general whose office handled the case. The cruelty of what Cornyn is doing is the subject of this essay, and the empirical record on retraumatization of child sexual-assault victims is the framework through which that cruelty becomes legible.

Begin with the family. The victim’s mother told the court that the plea Adam Hoffman accepted captured only a small portion of the abuse her son had suffered. She said so on the record. She did not pretend the plea was justice in any complete sense. She framed her acceptance of it as protective, not absolving, a choice to end the legal process and let her son begin healing rather than a pardon for what Hoffman had done. That distinction matters, because Cornyn and his political allies are now collapsing it. They are treating the family’s protective decision as if it were an exoneration of Hoffman, and treating the prosecutor’s acceptance of the family’s wishes as if it were proof of indifference to child sexual abuse. Neither claim survives contact with the actual record. The family wanted Hoffman to admit what he had done. He did. He pleaded guilty to indecent assault and to displaying harmful material to a minor, both Class A misdemeanors. He served 60 days, double what the original plea contemplated, after Visiting Judge Roy Sparkman increased the sentence. He surrendered his law license. He cannot apply for reinstatement for five years and almost certainly will never practice again. He is barred from contacting the victim. The mother and her son publicly stated that they forgave him. That is what closure looked like for this family, and it is what John Cornyn is now publicly second-guessing.

The prosecutor was Brenda Kaye Marmolejo Cantu, a 30-year trial lawyer with the Texas attorney general’s office. The first trial collapsed when defense counsel got the boy to acknowledge he may have exaggerated the number of abuse incidents in order to please a CPS investigator. Anyone who has worked with traumatized children understands what this means. It does not mean the abuse did not happen. Hoffman has now admitted, on the record, that it did. It means that a frightened 10-year-old, asked repeatedly by adults to count and recount and specify the worst things that ever happened to him, produced the kind of inconsistency that any developmental psychologist would predict. The defense exploited that inconsistency, as defense counsel are paid to do, and five jurors voted to acquit. When Cantu told the family she intended to retry, they balked. They did not want their son cross-examined a second time. Cantu, by her own admission to the family, had the legal authority to subpoena the boy and, if he refused to comply, to ask the court to hold him in civil contempt. That is the path Senator Cornyn is now suggesting Cantu should have taken. It would have meant a child being detained, or threatened with detention, until he agreed to face his abuser in court a second time. Cornyn is not saying this in those words, because no politician would survive saying it in those words, but that is the operational meaning of his criticism. There is no other lever Cantu could have pulled to force a retrial over the family’s objection.

The empirical literature is unambiguous on what that second trial would have cost the boy. Goodman et al., writing in 1992 in the Monographs of the Society for Research in Child Development, tracked child sexual-assault victims who testified against matched victims who did not. Quas et al., in a 12-year follow-up published in the same Monographs in 2005, returned to those same children and measured their adult outcomes. The single strongest predictor of long-term harm, measurable as elevated PTSD, depression, and behavioral disturbance more than a decade after the assault, was testifying multiple times. Direct face-to-face confrontation with the defendant worsened outcomes independently. Harsh cross-examination worsened outcomes independently. Weak corroboration, which is precisely the situation a hung jury reflects, worsened outcomes independently. Henry, writing in 1997, and Runyan and colleagues, writing in 1988, named the broader phenomenon system-induced trauma, the legal process becoming a second injury layered atop the first. Holmes and Slap in 1998, Romano and De Luca in 2001, and Dube and colleagues in 2005 specifically examined male victims aged 10 to 14 and found that the disclosure-related shame variables most predictive of poor outcomes in boys are precisely the variables that public testimony amplifies. Cornyn is asking, in effect, that the prosecutor and the judicial system override the family’s protective judgment in order to subject this particular boy, in this particular age range, with this particular trauma profile, to the single highest-risk intervention the empirical literature has identified. He is doing so without engaging the literature, without engaging the family, and without acknowledging that the entire reason 18 U.S.C. § 3509 exists, that Maryland v. Craig, 497 U.S. 836 (1990), exists, that the ABA Criminal Justice Standards 3-1.2 and 3-3.2 exist, and that the National District Attorneys Association’s National Prosecution Standard 2-8.4 exists, is to spare children from exactly the kind of compulsion he now demands.

Consider what those legal authorities actually say, because they are not advisory in the loose sense, they are the prevailing professional standards governing prosecutors. The Crime Victims’ Rights Act, codified at 18 U.S.C. § 3771, gives the boy enforceable rights to dignity and fair treatment. The ABA standards instruct prosecutors to weigh victim impact in charging and trial decisions. The NDAA standards specifically direct prosecutors to avoid subpoenaing reluctant child sexual-assault victims absent extraordinary circumstances. Civil contempt against a child crime victim, particularly contempt that contemplates detention, runs into Eighth Amendment limits, juvenile-code prohibitions on coercive detention when alternatives exist, and a documented line of bar-discipline and civil-liability cases. The well-known Texas and Houston cases from 2007 and 2015, in which prosecutors were sanctioned or sued for coercing child witnesses, exist precisely because the profession has decided this is not how civilized prosecution works. Cantu followed the standards. She also, it should be noted, made the strategically sound choice. A coerced, frozen, sobbing, or recanting child on the stand at a second trial does not produce a conviction. It produces another hung jury, or worse, an acquittal that frees Hoffman with a not-guilty verdict on his record and bars further prosecution under double jeopardy. The plea Cantu secured ended with Hoffman a convicted offender, jailed, disbarred, and permanently constrained. The retrial Cornyn demands would have ended, with substantial probability, with Hoffman free and the boy retraumatized for nothing.

So what is Cornyn actually doing? He is using a child’s suffering as a campaign weapon against Ken Paxton. The Texas Attorney General’s office has roughly 750 attorneys handling something on the order of 30,000 cases per year. No general counsel, no chief executive of any organization of comparable scale, is briefed on the line-level decisions of every assistant attorney handling a hung-jury retrial decision in McLennan County. Paxton learned about the Hoffman matter when Cornyn made it a political issue, not before. The chain of imputation Cornyn is asking voters to accept runs as follows. Cantu accepted a plea the family wanted. Cantu therefore protects pedophiles. Cantu works in an office Paxton runs. Paxton therefore protects pedophiles. The first link is false on the empirical record. The second link does not follow from the first. The third link is true but irrelevant. The fourth link is the entire purpose of the chain, because Cornyn is losing the runoff and needs an attack on Paxton’s character that voters will absorb without examining its premises. The boy and his family are the instrument by which that attack is delivered. They did not consent to that role. They are now watching it unfold on cable news.

The cruelty here compounds. The first wound was the abuse itself. The second wound was the original trial, with its cross-examination and its hung jury. The third wound was the reality, communicated to the family by an experienced prosecutor, that the only path to a conviction at retrial would have run through their son being subpoenaed and potentially detained until he testified against his will. The family chose to spare him that wound by accepting the plea. The fourth wound, the one Cornyn is now inflicting, is the public reframing of that protective choice as a moral failure. The mother is now reading commentators argue that her decision to protect her son makes her complicit in future crimes Hoffman has not committed and may never commit. She is reading that her son, by refusing to undergo a second cross-examination, is letting his abuser get away with rape. Survivor’s guilt is not a metaphor. It is a documented clinical phenomenon, and it is precisely the kind of psychological pressure that the empirical literature identifies as a long-term harm vector for child sexual-assault victims. Cornyn and his allies are, with their public commentary, generating exactly the cognitive content that the literature warns against. They are doing so in service of a U.S. Senate campaign.

There is a final point worth making, and it concerns the structural asymmetry between Cornyn’s position and the family’s position. The family, by accepting the plea, made a decision that they have to live with for the rest of their lives. They will second-guess it on hard nights. They will wonder whether they should have pushed for the retrial. They will read every story about a future Hoffman victim, if one ever exists, and ask themselves whether they could have prevented it. That weight is theirs to carry, and it is unimaginably heavy. Cornyn, by criticizing the plea, makes a decision he will carry for exactly as long as the news cycle lasts. If he wins the runoff, the criticism vanishes from his public record within a week. If he loses, it vanishes faster. The asymmetry of consequences is total. The family bears the lifelong cost of a decision Cornyn is exploiting for short-term political gain. That is not a misjudgment by Cornyn. It is a choice. He has weighed the family’s suffering against the marginal utility of an attack line on Paxton, and he has decided the attack line is worth more. That is the sentence that should follow him for the rest of his career, because it is the sentence that accurately describes what he has done.

The empirical record favors the family. The legal standards favor the family. The strategic logic of trial practice favors the family. The moral framework of victim’s rights, codified at federal law, favors the family. The 30-year prosecutor who handled the case favors the family. The judge who reviewed the plea, and who increased its sentence, accepted the family’s framing. The only person who disagrees with the family is the senator who needs to win a runoff and who has decided that this child’s suffering is a usable political resource. Texas Republicans are being asked, on May 26, 2026, to decide between Ken Paxton and John Cornyn. The Hoffman case, as Cornyn has chosen to deploy it, is the clearest possible window into how each man weighs the welfare of an actual Texas child against his own political position. Paxton’s office, through Brenda Kaye Marmolejo Cantu, did what 30 years of trial experience and the prevailing professional standards counseled. Cornyn is on television demanding that the boy be put back on the stand. Voters can decide which of those two postures more closely resembles the conservative tradition’s commitment to the protection of children, the rule of law, and the dignity of the individual person. The answer is not difficult. It is, in fact, the easiest moral question in the entire 2026 Texas Senate runoff, and Cornyn has, with extraordinary efficiency, answered it himself.

If you enjoy my work, please subscribe: https://x.com/amuse.

As Sponsored by the John Milton Freedom Foundation, a nonprofit dedicated to helping independent journalists overcome formidable challenges in today’s media landscape and bring crucial stories to you.

READ NEXT: Person Shot By Secret Service Near White House