Bryan Caplan  

Designer Babies Are Nothing to Fear: A Reply to Dan Klein

Politi-Obfuscate... Great Rules for Discussion...
I'm delighted that Dan Klein is engaging my defense of designer babies.  Here's my reply, point-by-point.  Dan's in blockquotes, I'm not.
The idea that technology will enable designer babies fills me with apprehension. My apprehension over designer babies would be great even on two unrealistic assumptions: (1) That the government never got involved in it (specifically, never made use of it or influenced the use of it), and (2) That the technological development did not lead to reductions in liberty, reductions prompted by results of the development.

Designer babies would attenuate coherence with the past: One hundred years hence, people would say, "When you watch him on the old videos, he may not look like much, but back in the old times, Mike Tyson was considered a pretty menacing fighter." People in the new times would not know our sense of standard. And they would have difficulty knowing it. Everything preceding the break, from Achilles to Rafael Nadal, would be foreign and unintelligible.

You could say exactly the same about all the economic growth that happened since 1900.  It has undeniably and dramatically "attenuated coherence with the past."  I'm tempted to say "So what?," but what I really think is, "Good riddance."  The world's improved beyond anything my great-grandparents could have imagined.  I hope my ancestors would have had the good sense and benevolence to be happy for me.  I'm happy for my super-descendents already.

People would lose a sense of historical coherence, a very important dimension in human meaning. But even within their time, across arenas, people would lose a sense of standard.
I'm a lifelong history buff.  The main thing I learn from history is that the past was awful, and the present far from satisfactory.  A future society of designer babies should be more able to appreciate this lesson because they'll be beyond so many of our failures. 

In any case, most people currently have almost no sense of "historical coherence," so they won't be losing much. 
People would diversify in extremes, to the point of regarding those in other arenas of activity (the musicians, the athletes, the scholars, the thinkers) almost as separate species, literally a specialized breed, and be little able and little interested in trying to relate to them. "Am I supposed to applaud?" "Am I supposed to try to remember his name?"
We're already grateful for the amazing fruits of specialization that we already enjoy.  I see no reason not to embrace further specialization. 

Whatever sense of standard people did have, they would expect it to dissolve quickly. So the breakdown in historical coherence would be in relation to both the past and the future. Every generation would bring drastic changes in standards, changes that were unforeseen and yet unsurprising.

Again, this applies equally to rapid economic growth.  Think about the breakdown of "historical coherence" between Maoist communes in 1960 and modern China.  Good riddance.

I've used examples from sports, but I think the troubles pretty well carry over to broader areas of life, including even wisdom and virtue. A main point of Polanyi's book The Study of Man is that reverence, that necessary instrument for perceiving greatness, is especially applicable in the pursuit of wisdom and virtue. Those, too, might have some genetic basis. But whereas we would still know which ball players hit the most home runs, here we might have even greater difficulty recognizing the standouts. And inasmuch as we did recognize the standouts, or thought we did, we might regard them as we regard the new home run champions: more as a specialized breed than as exemplars. Would I ever have bothered to dwell in Michael Polanyi's thought if I knew he were a designer baby?

The Dan Klein I know would focus on the quality of Polanyi's arguments, not his origin story.  That's what we should all do.

Adam Smith held that all moral approval relates to a sympathy; the sympathy ratifies or underwrites the approval. Suppose, on Smith's authority, that the principle is sound. It would be just as sound in a world of designer babies. But in such a world, the sympathies themselves would be terribly attenuated, and hence also the moral approvals underwritten. I think our moral confusion would grow more confused; I suspect that the result would be moral and spiritual life that is shallower, not deeper.

Dan, do you really think parents are going to select against genes for sympathy?  That's very hard to believe.  Kindness is one of the main things parents try to instill in their kids.  If anything, we should expect designer babies to feel more sympathy, not less.

Now, here are some reasons why the two assumptions are not realistic (and why my apprehensions are greater still): If political disposition in an individual has some genetic basis, then, just as governments got involved in schooling, governments would likely get involved in designing babies.
Possibly, but I'm not worried.  In Western societies, controlling reproductive choice is widely seen as totalitarian.  Who today does not recoil in the face of the Supreme Court's notorious 1927 decision to allow mandatory sterilization?  I am however worried that Western societies will deprive individuals of the right to use reproductive technology as they see fit, preemptively killing off an promising source of human progress.
As for other repercussions on liberty, consider these: (1) Designer babies would devastate social coherence, connectedness, and personal meaning as generated by voluntary, non-governmental affairs, and consequently, in demanding them, people will look ever more to governmentalization;
I can imagine designer babies would lead to marginally worse outcomes along these lines.  But why on earth would you expect designer babies to "devastate" anything? 

Though to be honest, I hope Dan's right.  In my view, existing levels of "social coherence" and "connectedness" are dangerously high, the cause of most of man's inhumanity to man.  See Dan's great work on "the people's romance." 
(2) If designing your baby is expensive, "the rich get richer" and "level the playing field" will be louder than ever.
The same goes for any high-cost novelty.  Fortunately, the market usually outmaneuvers populist bellyaching long enough to turn novel luxuries into affordable conveniences.
Liberalism brought rapid cultural change (see the first of the two trends I present in this 17-min video.) I detect, especially in libertarians, a denial regarding the downside of rapid change - and I mean a downside even apart from resultant greater governmentalization.

But don't get me wrong about libertarians; they are not unique in letting their ideological commitments distort their interpretations and judgments.

Count me a libertarian "denier."  Downsides notwithstanding, technological and economic progress are great.  To even selectively rebut the massive presumption in favor of progress you need to point to something like Hiroshima in 1945, not intangible worries about historical coherence.

COMMENTS (26 to date)
David R. Henderson writes:

I told Dan Klein in an e-mail this morning that the further I get away from his argument, the less convinced I am. Bryan has stated my objections plus others I didn’t think of.

Brian martens writes:

What If the parameters programmed for a potential human are not the ones they are satisfied with? Why should you choose their fate? This debate, in my opinion, should not be about society as a whole, but individuals and freedom. It does not seem right or just to design your child to your own liking and not theirs. Whether or not you agree that they could possibly have their own liking while unconveived is not debatable, they can't. But once fully grown, any depression over their design is entirely the parents fault.

khodge writes:

"The amazing fruits of specialization" is one of the areas that science fiction writers are especially wary of: (1) to specialize people means to create a stratified society with the elites making the decisions; (2) what exactly do we do with the horse and buggy people when the automobile people are designed?

Finally, to expand on my (1), how, exactly, are you going to control the elites? We already have them trying to dictate the economy...just wait until they get their hands on reproduction.

Alex writes:

Ok, but what about if people wanted to abuse this whole thing, like having kids with wings or something crazy like that?

There are some things to be concerned about.

Pajser writes:

Imagine that you design your son, and doctor asks you "how aggressive you want him to be."

If you chose median or below, your son will be "beta male". He might be successful in his profession, but it is less likely that he will be the boss, and he'll have smaller chance to seduce the most attractive girls.

Most people wouldn't chose "very aggressive" either, but I believe that almost everyone would chose "slightly or moderately more aggressive than median."

Tragedy of the commons, isn't it?

Nic writes:

First, a short comment on Mr Caplan's first article:

Similarly, once we allow government to genetically screen out violent temperaments, it will be tempting to go further and screen for conformity.

Why only applying this to the government? I don't see why parents won't be tempting to screen for conformity as well. On the aggregate level you might get diversification but individually I find it tragic that we may get moral duplicates of our parents - but well designed.

Although I think it is important to address the concerns, I find it hard to believe that it will change anything. If we will be able to create perfect designer babies in the future, one will start doing it and by becoming dominant over the ones who hesitated, they will eventually catch up.
My greatest concern is that we will loose our purpose and fascination of life. It will be unprecedented difficult for our perfect descendants to distinguish and be proud of themselves.

Henry writes:
The same goes for any high-cost novelty. Fortunately, the market usually outmaneuvers populist bellyaching long enough to turn novel luxuries into affordable conveniences.

Indeed, I struggle to think of a single example of a new high-cost technology that gave a large, enduring edge to rich people.

When only rich people could afford cars or cell phones, both sucked by today's standards. Nowadays rich people can afford very good cars or cell phones, but even globally poor people can buy cars or phones that are 90% as good at the basics of what a car or phone should do.

Jennifer Burden writes:

I think you need to read: Unnatural Selection: Choosing Boys Over Girls, and the Consequences of a World Full of Men by Mara Hvistendahl, to see some of the negative consequences we ALREADY have from designer babies.

Mike D writes:
It does not seem right or just to design your child to your own liking and not theirs. Whether or not you agree that they could possibly have their own liking while unconveived is not debatable, they can't.

You could say the same thing regarding virtually any decision made by parents of young children.

"Why should you get to decide which preschool to send Timmy to? Shouldn't it be up to him?"

The prevailing social norm seems to be that parents are given autonomy in making decisions for their children, except in instances in which the decisions they are making are blatantly against the interests of the child.

I think the same standard works in the case of designer babies.

In other words, designing a baby to have high intelligence is fine. Designing a baby to have no arms is not fine.

Bostonian writes:

'Who today does not recoil in the face of the Supreme Court's notorious 1927 decision to allow mandatory sterilization?'

If the prospective parent is mentally retarded, and if the offspring is likely to be so, I think lots of people would support mandatory sterilization.

Ike writes:

The Mike Tyson example is an interesting one. I imagine Floyd Mayweather will want to have a kid who is an aggressive boxer like himself, and many of his fans will want the same for their kids. Maybe it actually is a good thing if rich weirdos are the first ones to get the tech; I imagine they will make better choices than randomly chosen members of the population.

It's not clear to me that governments are *that* much of an issue, since the impact of any changes in gene regulations would only be felt 20 years out. What agents are operating on such a long time horizon?

Regarding "beta males", I imagine that many parents having daughters will choose to make them relatively non-promiscuous, since that is socially desirable. That suggests that the slow mating strategy of "beta males" will benefit at the expense of more aggressive strategies. (What socially desirable trait should we be most worried about proliferating?)

Maybe our best hope is that the industry structure producing designer babies turns out to be a monopoly or oligopolistic competition, and the players in that industry place restrictions on the designs people can make (e.g. no psychopaths) in order to maintain a good image with the public. Well, I guess for the people who *really* want aggressive kids there will always be firms to supply that demand :/

Maximum Liberty writes:

One of the greatest favors my father ever did for me was to cure me of my romanticism about the past. He pointed out that for 99.99% of people in the middle ages, life was a miserable toil until an early death. He pointed out that, given his choice of all historical times to live in, he'd choose right now because people generally were so much better off than any time ever. Your post struck me in the same way.

Max L.

Hazel Meade writes:

There are a couple of good posts on this subject over at Marginal Revolution.

It seems as if most of the fears are almost entirely based on speculation drawn from science fiction novels, in which there is usually some sort of totalitarian state, not on putting the technology in the hands of parents. I also don't think the fears of the rich making their kids a genetically superior social class are very well thought through.

As Neem points out at Marginal Revolution, it's a lot harder to engineer for traits like intelligence, which involve hundreds of genes, than it is to correct specific defects responsible for genetic disorders. Second there are risks of creating other defects, which itself should be a disincentive to parents to go beyond single-gene corrections. Third, you can only do it via IVF which is expensive and painful to start with.

I think the fears of people creating a genetically engineered race of superhumans are wildly overblown. The technology itself hasn't been improved to the point where one can precisely target a single gene, much less a group of hundreds controlling something like intelligence or heirght. And the disincentives to parents to do more than just the minimum are high, both in terms of cost and risk.

Plus all of these traits already exist in the human gene pool. Nobody knows how to make a superhuman, yet. We're just talking about modifying a gene to change it from allele A to allele B, both of which are already present in the gene pool. We're not talking about transgenic humans, we're talking about optimizing combinations of alleles that already exist.

NZ writes:

People generally react to the prospect of designer babies with skepticism, an earnest attempt to weigh pros and cons, and a sober consideration of what the new technology may cause us to lose.

There's a bit of repulsion as well, typically grounded in a certain nostalgia for the past or even the present.

Personally, I'm not terribly concerned about radical change brought about by designer babies. Realistically in the short- and medium-term, I think the technology will allow people to maximize or minimize the odds of certain already-existent genes manifesting as traits--specifically, the most visible and genetically simple ones. Plus, the people using the technology will be mostly people who are already very careful and selective about who they mate with, and who already are likely to lavish lots of attention and care upon their child.

What really interests me is how we can get people to think about other prospective technology the same way. Nobody had this reaction to social media, but I'd say social media has had a much more widespread, devastating, and depressing impact upon both the individual mind of the social media user and upon society as a whole.

Hazel Meade writes:

Ok, but what about if people wanted to abuse this whole thing, like having kids with wings or something crazy like that?

Yeah, we really need to base this debate in science and not science fiction.

As someone with a layman's knowledge of biotechnology, I'm just going to say right now that that's completely impossible. Not only with CRISPR, but ever.

There are a lengthy list of reasons, starting with that making an entire physiological structure like wings involves more than our current theroetical knowledge of genetics permits. What we do know, at least is that wings are arms, so you'd really be trying to make a deformed baby with no arms and wings instead. Physiologically birds also need extremely powerful arms and small bodies, so the wings wouldn't even be functional unless you modified a lot more than the arms.
But the even bigger reason is that genes don't program for things like arms vs. wings, they program for an developmental growth pattern that unfolds in a particular way. Any modification you make to that program has enormous and uncontrollable downstream effects. Anything that effects something as large as arms are almost certain to result in nothing but a dead fetus.
CRISPR can only make specific edits. We're nowhere near being able to write DNA code like we were programming in C.

If were making policy choices based on totally hypothetical speculations like this, we should stop.

NZ writes:

Other new or upcoming technologies worthy of similar scrutiny:

  • "Smart tech" in cars (blind-spot alerts, driver fatigue sensors, etc.)
  • Driverless cars
  • "Smart" houses/appliances
  • Augmented reality
  • "Wearables" (wirelessly connected sensors)
  • Various types of aerial drones--especially those capable of surveillance
  • "Smart tech" for kids
  • Uber/other types of jitney transportation
  • AirBnB and other types of room-letting services
  • Advanced realtime facial recognition software (e.g. in marketing)
  • And, to Hazel Meade's comment above, wearables designed to enhance human performance (a mechanical pair of wings, for example. Or for a more realistic example that's currently available, dog ears that move according to your facial expression)

[Typo fixed: "David Henderson's comment above" changed to "Hazel Meade's comment above" by commenter request--Econlib Ed.]

Hazel Meade writes:

Nobody had this reaction to social media, but I'd say social media has had a much more widespread, devastating, and depressing impact upon both the individual mind of the social media user and upon society as a whole.

Your comment is spot on in a lot of ways, but I really have to give my approval to this. Most social media websites are pretty much explicitly design to manipulate people and get them addicted via social status signalling. I really think this has messed with a lot of people's heads in some very negative ways for society - starting with the recent trend of using twitter to socially ostracize individuals as part of a mass execercise in moral signalling.

Bostonian writes:

We could "genetically engineer" for IQ right now, by having personal ads list SAT and ACT scores, since standardized test scores are highly correlated with IQ. I wonder why it is considered normal to mention to a date your educational background but not your test scores.

Someone who got a BA from Harvard is on average more intelligent than someone who graduated from UMass Amherst, and both are more intelligent than non college graduates on average. There is assortative mating by degree and perhaps by prestige of degree. But somone may have gone to a state school rather an Ivy for reasons other than IQ.

NZ writes:

Thanks, Hazel.

The big lesson is that when it comes to new technology it's good to take concerns like those of Klein's seriously, maybe even to give them the benefit of the doubt. (It's the same as we do for new ideas, isn't it?)

I imagine that if 10-20 years ago people had expressed similar concerns about the advent of social media, then you'd have heard the same response from tech bravoists, including a lot of Bad-Old-Days "Good riddance" libertarians like Kaplan, of "The past was awful, and the present far from satisfactory. A future society of [social media] should be more able to appreciate this lesson because they'll be beyond so many of our failures."

Who knows, maybe that's exactly what happened. I wasn't reading blogs back then. Those were the days.

NZ writes:


People think differently about something when there's a number attached to it. There might be more of a tendency to obsess over the number, to let it become all-consuming, and to get lost in the number comparison rather than consider the wider implications.

For example, having gone to Harvard doesn't just mean you're smart; it also means you probably know lots of other Harvard people, many of whom might have kids who can provide great peers to your prospective kids. It also means your kid will one day have a legacy advantage should he decide to apply to Harvard. Just knowing that someone got a 1600 on their SATs doesn't tell you all that, and in an era of rampant test-gaming it tells you less still.

I also think there's a vague awareness of this fact, which is why people consider it normal to share educational background with a date right up front, but not test scores.

Xenophon writes:


Your comment perfectly illustrates one of the dangers of taking research too literally.

TTBoMK, the research you to which you sort-of allude utterly fails to distinguish between (for example) “aggressive,” “assertive,” and “justifiably self-confident”, any of which could make the difference between “alpha” and “beta” status. But the very real difference between these descriptions has huge effect on the implications of the rest of your comment.

Zkidron writes:

I am worried that we, a product of 2 billion years of evolutionary trial, success and error, now want to take the error part out of the equation. I have an inkling nature gained as much by the "errors" along the way as by the successes.

In a sense it's akin to trying to establish a market economy while minimizing loss and maximizing profit, all by rigging the laws of economy through "intelligent design".

Genetic engineering will in the long term reduce genetic diversity and thus possible future evolutionary breakthroughs.

I find it a bit odd that economists such as Caplan who would not hesitate to criticize intents to design and direct a complex system such as an economy can still think positive an intent to design and direct a no less complex system, the human genome.

Yes, I'm aware for the moment we only posses enough knowledge for minimal genetic "edits" useful only for correcting specific genetic defects in the best case scenario. But that won't be the case for long and the long term effects should be evaluated.

Prakash writes:

I would re-iterate the point that sex selection has issues right now, in China and India (even though it is banned in India, it prevails. Those sex ratio numbers are not natural by any means).

Someone may counter that western cultures are unlikely to make that mistake. But it's precisely your current familiarity with that culture that allows you to say that, the very familiarity that is expected to be extinguished when the genetically modified take over.

You could make a generalised libertarian transhumanist statement and I would respect that. That is a principled position. But argument arising from current interpretations of western culture seems a bit like trying to have your cake and eat it too.

Hazel Meade writes:

I find it a bit odd that economists such as Caplan who would not hesitate to criticize intents to design and direct a complex system such as an economy can still think positive an intent to design and direct a no less complex system, the human genome.

Except that's not what anyone is proposing nor is it likely technically feasible (precisely because it is too complex a system).

They're talking about single-gene edits, not "engineering" major modifications.

We're not anywhere close to having the theoretical knowledge necessary to attempt to "design" a genome. Any attempts to do such a thing would almost certainly result in non-viable embryos.

To use your analogy, this is like saying that modifying a single line in the federal register is equivalent to "engineering" the entire economy.

ams writes:

This isn't specific to genetic engineering, but I've been thinking:

People have a natural tendency to want certain things out of life: A lot of the things we find desirable or goals derive from needing them to keep ourselves and our children healthy and alive. In markets, we organize ourselves to provide each other with things we want as efficiently as we can within our means.

Okay, now what happens if some of these things on the market include the ability to modify the sort of things we want? (I myself have often wanted to want certain sorts of things) If genetically engineered generation N+10 has a completely different set of fundamental motives than generation N on account of the way they've been modified (probably with arms-racing thrown in as a driver). Presumably N+10 achieving what N+9 desired within the internal logic of their value system, but how trustworthy is it?

The monkey-cocaine experiment comes to mind. Prior to pushing the button, the monkey is motivated by all sorts of monkey things. After pushing the button enough times, the monkey is entirely motivated to just keep pushing the button. By a completely utilitarian understanding of the goodness of an outcome, post-button-monkey is getting what post-button-monkey desperately wants, even though pre-button-monkey would find his motives incomprehensible.

I dunno - seems to be the potential for instability/danger there. Not advocating for not having the choice though. (Yet)

ams writes:

I suppose as long as generation N+10 is *functional* and self supporting within the world they exist within, then that could be one external measure of how good the outcome is for generation N+10.

(After all, we live radically different lives than 10 generations ago, and we seem to do all right within the context of our radically different value system.)

How do you distinguish between a scenario where the goals people are pursuing are meaningful, and a situation where the goals they are pursuing they are pursuing because they've gotten mode-locked somehow? It seems like you would have to appeal to something else besides how it evaluates in terms of local utility functions: How functional it is in terms of an environment maybe?

Comments for this entry have been closed
Return to top