dissenting:
I respectfully dissent because I believe any error in the jury-selection process was harmless.
I
The defendant Cameron Dean Bates went to trial on child-pornography charges. The evidence of guilt was overwhelming. Mr. Bates says with more time-he could have developed evidence support*892ing his defense, but he still has tendered none.
The undisputed, forensic, objectively verifiable evidence included the following. Mr.- Bates’s laptop computer contained a large volume of child pornography. The laptop also contained numerous images of Mr. Bates nude or engaged in sexual conduct with men as well as sexually oriented craigslist advertisements that Mr. Bates placed. These materials were not password-protected. The laptop also contained legitimate, unobjectionable materials readily linked to Mr. Bates, .including, for example, word-processing documents that he authored. Child-pornography images were accessed very close in time to when work was done on some of the materials readily linked to Mr. Bates. The compelling inference was that Mr. Bates not only worked on the materials readily linked to him but also accessed the child pornography.
There was no forensic, objectively verifiable evidence that anyone else ever used the laptop. There were other computers in Mr. Bates’s residence; there was no apparent reason why Mr. Bates would make the laptop available to others. Family members said, though, that Mr. Bates left the laptop at the residence, readily accessible to family members and visitors, and that some used it.
The contested issue at trial was whether Mr. Bates accessed the child-pornography images, as the government contended, or someone else did it, as Mr. Bates contended. Mr. Bates pointed to a friend of his son as the likely culprit. The images of the defendant engaged in sexual conduct with men were admissible to show the unlikelihood that Mr. Bates would allow others — including his son’s friend — to use the laptop. The government had every right to prove those images.
Not surprisingly, the jury found Mr. Bates guilty. On this evidence, any reasonable jury would have reached the same conclusion.
II
Mr. Bates submitted in advance a comprehensive list of proposed voir dire questions. The list did not include any questions dealing with bias against gays or relating to adult sexual conduct. But during the jury-selection process itself, at sidebar, Mr. Bates’s attorney asked the judge to inquire on the subject. The attorney’s entire request was this: “I think you should be asking some questions relative to homosexuality and preferences of people and so forth.” The judge said he was not going to get into the subject during jury selection, that he did not know whether he was going to let in evidence on the subject, and that he would weigh probative value against prejudicial impact when the evidence was offered. Mr. Bates’s attorney said, “That’s all I asked.” If not acquiescence in the judge’s decision not to ask questions on this subject during the jury-selection process, this was surely close.
Whether the judge abused his discretion — in ' light of what the judge knew about the case and what the parties argued at the time — could be debated. It of course would have been preferable to ferret out any bias. But ferreting out bias is not easy; a simple question or two might not have made any difference. And questions on this subject might have adversely affected Mr. Bates, highlighting the issue of sexual conduct with men right from the outset. Not getting into a tangential matter during jury selection — thus leaving it to the attorneys to introduce the case for themselves during their opening statements — is ordinarily a reasonable approach. Mr. Bates did not suggest specific *893questions that would have addressed bias without getting into the facts of the case in a manner that would have harmed, not helped, Mr. Bates.
The judge of course could have addressed the subject of bias without prejudicing Mr. Bates. This could have been done, for example, through individual voir dire that could have been conducted quickly at side bar and that could have addressed attitudes about child pornographers as well as adult sexual conduct, avoiding any statement about what Mr. Bates did or did not do. But Mr. Bates never asked the judge to do anything of this kind. In the absence of any more specific request, it is not at all clear that the judge abused his discretion.
Ill
If, as the majority holds, the judge abused his discretion, the case turns on whether the error was harmless. The issue is this: is there a “reasonable possibility that [sexual-conduct-based] prejudice might have influenced the jury”? United States v. Groce, 682 F.2d 1359, 1362 (11th Cir.1982). The majority recognizes this standard and suggests that whether prejudice “might have influenced the jury” means the same thing as whether prejudice “might have contributed to the conviction.” See majority opinion at 888 n. 4.
The strength of the evidence is part of the analysis. There is a reasonable possibility that prejudice against scientists might influence one’s view on global climate change. But there is not a reasonable possibility that prejudice against scientists might influence one’s view on whether the sun rises in the east. This is closer to a sun-rises-in-the-east case.
The judge’s limiting instruction also is part of the analysis. The judge told the jury to consider the evidence of Mr. Bates’s sexual conduct with other men only on the issue of control of the laptop. The majority notes that the limiting instruction did not go as far as it could have, but Mr. Bates did not object to the instruction as given or ask for anything more. The judge also instructed the jury to decide the case without sympathy or prejudice for or against the government or Mr. Bates, to rely only on the evidence, and to decide only whether Mr. Bates was guilty of the specific crimes charged in the indictment. Again Mr. Bates did not object or ask for any further instruction on these matters. There is no reason to believe the jury failed to follow the instructions.
In light of the overwhelming evidence of guilt, and in the absence of any evidence of any juror’s actual bias, I would hold that any error in the jury-selection process was harmless. There is no basis at all for any assertion that the jury convicted Mr. Bates because of bias against gays rather than based on the overwhelming evidence that he committed the child-pornography crimes with which he was charged.
IV
I acknowledge, though, that the harmless-error issue is reasonably debatable. Bias against those who engage in same-sex conduct exists. And the majority’s position that any error was not harmless draws strength from the government’s conduct at trial. The government repeatedly asked questions on this subject to witnesses who knew nothing about Mr. Bates’s sexual encounters with men or the images on the computer. As the government knew before it asked the questions, the witnesses could give no testimony on this subject having legitimate 'evidentiary value. The government was plainly angling for prejudice. At least in hindsight, this makes more problematic the failure to address bias during jury selection.
*894One is left asking why, if the evidence of guilt was as clear as I believe it was, the government asked improper, prejudicial questions. A possible inference is that the government thought a conviction was not certain. A possible inference is that the government thought at least some jurors were biased and that appealing to that bias would help bring about a conviction. Why else would the government do it?
I am left in the uncomfortable position of concluding the government was wrong — that it didn’t need the prejudicial impact it improperly pursued. It is with no enthusiasm that I dissent.