Millionaire Robert Graham’s Repository for Germinal Choice (1980-1999) sperm bank was founded as a form of positive eugenics in order to encourage sperm donation by gifted men (initially Nobelists) for use in the nascent field of artificial insemination. Launched to instant infamy, it turned out to have actually struck a major chord among women seeking sperm, who were generally treated extremely shabbily by the medical establishment which when doing as it pleased, casually chose donors largely at random and denied the women any kind of choice or information about the donor (Plotz notes the first recorded case of artificial insemination involved abruptly chloroforming the woman and using a random medical student). However, it encountered perennial troubles in obtaining sufficient supplies, as artificial insemination (not necessarily/usually IVF, as I assumed for most of the book until I finally realized my mistake) used up large quantities of semen before a successful pregnancy, so the lack of Nobelist participants (between the rigorous medical testing and the notoriety) immediately forced a switch to less distinguished donors; further, fees charged to women never came close to covering the operating expenses of recruiting those donors and schlepping all the semen around, even as other sperm banks adopted the Repository’s innovation of stringent health examinations & forcing Graham to sustain the Repository himself, and while he arranged for millionaire Floyd Kimble to take over funding the Repository when he died, that millionaire then soon died himself without having made any further provisions! Graham’s family was happy to see the sperm bank die, and that was that.

Around 2000, journalist David Plotz began a 13-part Slate investigative report describing the positive eugenics background, history of the sperm bank, and trying to find donors/mothers/offspring - succeeding in reaching a small fraction of them. The online series includes some of their personal reactions to their experience, beliefs about the harm, some of them being reconnected with each other, descriptions of their current circumstances etc.

The first question about this book is, is it worth reading if you’ve already read the Slate articles and are interested in learning more? Yes. The background on Graham, Shockley, and modern sperm banking is much more extensive in the book, and it goes into substantially more detail about the donors/mothers/offspring. For example, the Slate series has one 2001 post focusing on “Donor White”, who had not been found by that point; but White showed up afterwards, was interviewed extensively by Plotz (much of the book is in the first-person), and interacted a great deal with Beth/Joy over the following years, all of which is in The Genius Factory but not the Slate articles. He also corrects/updates a number of assertions (eg how exactly the Repository closed, with the online version concluding vaguely that it must have shut down because Graham somehow just didn’t bother to put anything in his will and his relatives didn’t support it, while the book version fixes this by bringing in Kimble and explaining what went wrong; apparently none of these corrections have been added to the Slate versions, checking back).

It’s interesting seeing how disparate peoples’ reactions to the sperm bank are, ranging from (the proper) indifference to considerable curiosity to almost neurotic obsession. I also appreciated the book expanding on the descriptions of the offspring and their successes even in trying circumstances, and the modern sperm banking industry, which is hard to get a read on because it’s so private (eg Plotz quotes Repository staff noting that, as long suggested, prospective mothers value highly height and health; leafing through the catalogue, everyone is a positive eugenicist), and the issue of where the unrelated fathers stand (in a very difficult one, and at least for the women who contacted Plotz, in a generally untenable one, although he notes the selection bias). So I enjoyed much of the book and read it in one or two sittings.

Much of this is relevant to anyone thinking about the current prospects for embryo selection on traits. The estrangement of fathers emphasizes how naive it is to hope that merely offering some sperm of better genetic quality would be enough to encourage en masse usage: genetic relatedness is far too important to almost everyone, and giving up relatedness for better traits is inherently insulting to the cuckolded father; egg/sperm donors are always a last resort. (This is something the iterated embryo selection & genome synthesis approaches must grapple with; who will use your optimized eggs/sperms if it means the child will be 50% or 100% unrelated to the birth-parents? On the other hand, regular embryo selection & CRISPR preserve relatedness almost entirely.) The lure of greater intelligence turns out, surprisingly, to not matter as much to the mothers as does height/athleticism/health and avoiding below-average outcomes. So mothers prize physical attributes as much or more than mental ones, and are risk-averse; suggesting the importance of doing selection on multiple traits of which intelligence is only one and perhaps not even the most important one and of emphasizing that we have excellent height polygenic scores which right now would allow height increases of <4 inches, and of framing it in terms of reducing the chance of a low outcome rather than its equivalent increasing the mean.

What’s bad in the book? Plotz comes off, as a little snide & anti-intellectual; he seems to take an attitude in slightly disliking almost everyone in the book and it bleeds through unavoidably. He lacks any kind of sympathy. This slight disdain extends from the people to the core topics. Though he can’t deny the power of genetics when even the briefest meeting or description of the sperm donors shows their resemblance to their offspring, he is an orthodox liberal in doing his best to deny it. (Which lends some passages surreal qualities; having just described how successful a bunch of kids were or how they resemble their donor or conceded that intelligence is indeed heavily genetically influenced, he’ll then invoke the shared environment or epigenetics as the explanation of everything and move on. I am reminded of the story that Bertrand Russell, seated next to a Christian at dinner, asked what he thought would happen to him when he died: “Oh, well, I suppose I shall inherit eternal bliss, but I wish we wouldn’t talk about such an unpleasant topic.”)

He also makes a number of errors or questionable claims or perpetuates things he should know better. I noted down a few while reading:

He notes that the press hyped the Repository as the “genius factory” or the “Nobel Prize bank” or calls them “superbabies” or “genius babies”, and then he goes on and routinely uses those hyperbolic phrases himself and indicts the Repository as a failure for producing no geniuses, even after having correctly noted that the ‘genius babies’ would not have been anything of the sort because they would get only half their genes from the sperm donor:

What were the kids like? Had the genius genes created genius babies? Were Repository prodigies now skipping their way through America's best private schools, prepping for Harvard, intent on curing cancer and reinventing physics? Were there lots of little Shockleys out there, hot-wiring the latest Intel chips to work double time?...Graham thought his donors would supply a massive intelligence boost. In fact, the genetic improvement was probably minuscule. Nobel sperm would give modest odds of slightly better genes in the half share of chromosomes supplied by the father. And even then Graham would be operating on only the nature side of the equation: he had no control over nurture-schools, upbringing, parents. This was a formula for a B-plus student, not the “secular savior” Graham hoped to breed.

This is problematic because, aside from putting words in Graham's mouth who reasonably expected "a few more creative, intelligent people", he is judging the method fundamentally flawed when the results, as far as Plotz’s mini-census is able to uncover and Graham himself believed based on early reports (but was unable to confirm due to non-cooperation from the mothers), are consistent with what the simplest application of genetics would have predicted. At no point does Plotz figure out what the results should have been So I will do it for him. The adult heritability of IQ is ~0.8 now, increasing during childhood, because schools/upbringing/parents just don’t matter that much. The donors listed range in gifts, but an IQ of 130 seems like a reasonable guess given their general education and often scientific success (at least two donors should’ve been excluded by the Repository, but in both cases they are clearly well-above-average anyway). So they would be expected to yield a boost of +12 IQ points. The mothers themselves range from below average to perhaps 130s themselves, we’ll guesstimate 110. The offspring will be half-related to the donor and to the mothers; so their total expected adult IQ would be 30*0.8*0.5 + 10*0.8*0.5 = 16 or ~116 with the usual 15 SD; their childhood IQs would tend to be a bit lower. What would we expect from such a group? Well, we would expect them to do well in school, be healthy, athletic, a number of them at the top of their class and MENSA level - in short, we would expect what Plotz shows us, and we expect them to basically resemble a group of Ashkenazi children given mean Jewish IQs of ~110! (Incidentally, an especially high-scoring child, such as Doron Blake would be expected to regress back to 116 due to the major instability of childhood IQ; even if Doron Blake had scored at 160 or something, very early childhood IQ correlates r=0.5 or less with final adult IQ, so Blake would be expected to end up somewhere around (160-116) * 05 or 138 IQ.) A marginal +12 IQ points is no joke; that’s worth many thousands of dollars in annual income, increases the odds of graduating college, etc; and from an eugenic perspective, this is a gain that can accumulate over multiple generations. The world would look very different if each generation was 12 points smarter. (To put that in a global perspective, a mean of 12 points takes you from the UK or USA to somewhere like subsaharan Africa.)

Plotz’s timeline is hopelessly pessimistic when he writes

The Nobel sperm bank kids, I realized, were messengers from our future. We are on the brink of the age of genetic expectations. Soon-maybe not in 5 years, but probably in 50-fertility doctors will be able to identify and manipulate genes for “intelligence” and “beauty.”

Indeed, not in 5 years from 2005, but he knew full well that PGD existed in 2005 since he covers it in the book and was being actively developed, and had probably heard about the ‘Moore’s Law for sequencing’. It didn’t take 50 years, it took 8: the publication of Rietveld et al 2013 would make the identification & manipulation of intelligence genes possible, and PGD was already waiting for it. It can be done now if anyone wants to.

Describing Galton’s work:

Successful fathers had successful sons. This, Galton claimed, proved that God-given abilities were passed from one generation to the next. (It did not concern Galton that in Victorian England, advantages of birth, wealth, and education might have given the sons of famous men a career boost.)

Wrong. Galton was well aware of the issue and tried to figure out the effect of such environments, inventing the adoption study, and finding - exactly as subsequent studies using a variety of designs have also found - that the ‘advantages of birth, wealth, and education’ didn’t count for much. Sloppy axe-grinding.

On applications of eugenics:

The American eugenicists’ most important cause was sterilization. How they longed to cut! They thought practically everyone should get the knife: the “feebleminded,” alcoholics, epileptics, paupers, criminals, the insane, the weak, the deformed, the blind, the deaf, and the mute-and their extended families. Of course, most of the purportedly genetic ailments developed by eugenicists were not, in fact, genetic in origin.

Wrong. All of those are highly heritable and many genetic variants for them have been found, particularly alcoholism, insanity (presumably schizophrenia), and deafness. (Plotz’s arrogance is particularly offensive here as even in 2005, hundreds of deafness genes had been identified.)

Oddly, another trait that doctors sometimes tried to match was religion, as though it had some genetic component.

Religious attitudes are heritable.

On speed of eugenics:

And even if they had been genetic, sterilization would have been a hopelessly bad cure for them. It would have taken literally thousands of generations of mass sterilization to significantly reduce the incidence of genetic diseases. But eugenicists didn’t stop to do the math.

Likewise wrong. I have no idea where Plotz got this claim of ‘thousands of generations’ as he doesn’t cite it, but where to start… Non-disease traits respond extremely quickly to selection, which would justify eugenics on its own quite aside from diseases; the commoner diseases could be substantially decreased within a few generations (I calculated that after 20 generations, schizophrenia could be halved, which is more effective than any other anti-schizophrenia treatment currently in use…); while it might take ‘thousands of generations’ to completely wipe out a particular disease, that will be because it had already diminished to a great extent and as it becomes ‘harder’ to wipe out that becomes ever more unimportant; eugenicists did stop to do the math because eugenicists like R.A. Fisher invented the math.

the timeline of behavioral genetics is quite bizarre:

late 1970s. At the time, sperm collection was practically the only widely available fertility treatment that worked. Social science research was beginning to show that intelligence was at least partly heritable.

Well before then.
Plotz cites uncritically both empirically falsified Gardner’s “multiple intelligences” and epigenetics
many of Plotz’s criticisms make no sense or are self-contradictory; he lambasts the Repository for the idea of focusing on Nobels, and then writes “Graham wouldn’t have known what to do with an oddball like Einstein.” Um, no, I think Graham would’ve known exactly what to do with a Nobel Prize winner like Albert Einstein, since you just wrote an entire book on that topic.

a deeply disturbing anti-intellectualism trend surfaces in his descriptions of Shockley. I was particularly struck by

Shockley himself didn’t seem like much of a provocateur. He discussed incendiary topics in a bizarre manner - exactly as if he were summarizing the latest advances in semiconductor research. He was the iceman. He didn’t exude hatred for blacks - he didn’t have any. He didn’t exude sorrow - he didn’t have any of that, either. Shockley’s critics assumed that his racial anxiety stemmed from some personal experience, some deep trauma, but it probably didn’t. He had no particular feelings for blacks one way or another. He hardly knew any blacks. To him, his racial conclusions were simply the logical outcome of a train of thought. As far as he was concerned, once he started to address human quality, he would follow its logic wherever it took him. In his mind, his conclusions had nothing to do with any actual black person; he was simply making an irrefutable point.

One might think that in discussing a highly controversial and highly important topic, being dispassionate, having no personal grievances, and attempting to hew strictly to the science and logic would be laudable. Apparently not. Apparently if you care about it, you’re a racist; if you are scientific and unbiased, then you’re ‘bizarre’ and the ‘iceman’ and still a racist. This total lack of sympathy or interest in understanding Shockly’s points leads Plotz into another genetics blunder:

Shockley thought he could prove to blacks that whiteness led to intelligence. Shockley proposed to do this by measuring the percentage of “white” genes in blacks: he would show that the “whiter” the black person, the smarter he was. (Not that he had any real idea of how to test for “white” genes.) He asked NAACP leader Roger Wilkins to help him collect blood samples from members of the Congressional Black Caucus and other celebrated blacks, on the grounds that these accomplished people would surely prove to be significantly white. When Wilkins rejected him furiously, Shockley suggested that Stanford blood-test its five hundred black students. You can imagine how well that went over on campus.

Extracting racial ancestry and ‘white genes’ is hardly as difficult as Plotz makes it out to be, and Luigi Luca Cavalli-Sforza was busy doing just that at the time; ‘admixture studies’ have been extensively used throughout medicine to help pin down disease-causing variants which differ by race, and - just as Shockley proposed - have been used in the debate since then.

more overvaluation of shared-environment:

The more I thought about it, the less surprising the maternal resemblance seemed. Most of these children had been raised only by their mothers. Their “social fathers” tended to be emotionally distant, and their biological donor fathers were out of the picture. So of course they were tied tightly to their moms. The mothers were women anxious for children, so motivated that they had chosen a genius sperm bank. Not surprisingly, they had become driven mothers. They spent more time with their kids than most parents did, certainly more than I did with mine or than my wonderful parents had with me. Was it any wonder their children grew up to be like them? I got the feeling that Samantha could have taken sperm from the dumbest player on the NFL’s worst team and would still have raised a brilliant boy. Her good genes would have helped, but so would the stimulating world she created around her. Any child would have fallen under that spell.

Plotz ignores that he spends much more time with the mothers than the donors in his quest to rescue shared-environment.