Assorted links

1. Appreciation of John Dunkley.

2. Confidence pays off for pundits.

3. What to do if you find a hedgehog (non-ironic, and indeed it is from the British Hedgehog Preservation Society.  Furthermore the results apply even if you have not “found” a hedgehog but rather are “concerned” about one, the site recognizing appropriately that perhaps he or she has found you).

4. Chris Hughes and Google are giving away cash directly to the poor.

5. Should we weaken the employer mandate?

6. Academic disputes over “genius babies.”


#6 Just a little tought experiment: let's say these guys after a few years claim that they have found the intelligence gene and that they can implant it in fertilized eggs. Children will have a IQ level 5-15 point higher.

The only problem: all this shit is not falsifiable. And if it is, it will take a long long time to find about it. A guy can claim success and sell it to "rich" people. But it can be just the naked emperor story, one more time.

Just a methodological comment, I think the idea is that research might identify hundreds of markers associated with a trait, then they'd fertilize perhaps 20 eggs for a couple, grow them a little bit, screen them to find the "best" weighted across all these markers, and implant that one. Then you wind up with a baby two standard deviations above the parents in that trait. Or at least that's the idea. There's no gene splicing or advanced genetic engineering the way lay people like me think of it. It could conceivably work for lots of different traits that might have significant genetic components like IQ, height, attractiveness, sexuality, etc. Practically speaking I would worry that you'd need to understand these genes in detail, and in particular how they add up, in order to make this work. Correlation to the trait might not be enough. Further, you couldn't optimize for a bunch of things, as you'd be unlikely to get an embryo that's 2 sd from the mean in more than one trait unless you had a really large number of embryos to work with.

You'll see this work with lots of genetic diseases long before you see it done for more ordinary reasons.

Does "g" even exist, or is it just a statistical artifact? Somewhere else in Michigan...

Elsewhere in Michigan, Noah Smith discussed that and a critique of it here:

#2 - PK

This is much better.

"#2 - PK" - Good call, but I think you'd have to also blend in some of Dan Ariely's work on dishonesty to nail that one, as we saw once again in last weekend's exchange with Reinhart-Rogoff.

funny, I was just talking with someone today who noted that PK and R&R are all hedgehogs. ... I wonder what the bedraggled pundit equivalent of "meaty cat or dog food" is? (#2 + #3)

Tyler is the foxiest of them all.

Perhaps. I know plenty of foxes ... but what's sad is most of them get lazy, believe one of their stories, and go hedgehog. Btw I am happy you got my fox-hedgehog reference ... my original comment was not based on looks or palates.

"Their hypothesis: Pundits have a false sense of confidence because that's what the public, seeking to avoid the stress of uncertainty, craves."

The problem is us. And by "us" I mean you guys.

For some people, this should be their Trayvon moment (my last straw with the media) where they lose faith in everything they thought provided them certainty and they realize that they are set adrift and had better start rowing.

derp, derp. come on, Andrew'. the pundits are useful (and thus rewarded). sure some people crave the fake certainty or enjoy betting on the 'winner,' but a sea of know-it-all's is actually a good way to span the set of outcomes. even If you are more the mutual-fund-of-info type, you still benefit from some loud (opposing) pundits in your portfolio too.

I'm not talking about the people who cling to more than one know-it-all.

Two would be a huge improvement, as long as the second isn't just a coat tailer of the first.

I'm not that hard to please. Like Dilbert, I ask so little of life and boy do I get it.


That's a nice point, especially if the pre-implantation selection technique promises only a smallish effect, a kind of tailwind for your descendants. What can be checked via DNA testing is that the baby has whatever high-IQ-associated alleles the provider claims it has. With more work, someone can verify that this baby is plausibly a lot higher than expected from these parents in those alleles. But determining whether or not the kids that result from this technique are notably smarter could take a long time. I'm assuming we won't see randomized control trials here, but rather that some early adopters will have it done in hopes of getting smarter (healthier, taller, better-looking, more emotionally stable) kids.

Exactly, as an early adopter all you can do is contract an independent tester and make sure your baby has the gene tracers associated to intelligence as your original provider promised. Whatever that develops in adult life intelligence is unknown.

The problem is that our smug & anxious culture says that we already have lots of gifted children. So, if as a parent/China government you payed big bucks for having a smart child, I don't expect parents saying 20 years later "I was scammed, my child is just average".

You can do randomized control trials for high yield soja and maize in a few months, and then try again and again and again. Not for children. It may work, but prepare for a 100+ years research period.

#5 - Josh Barro links to a WSJ article with this:

"Officials in several states are seeking to stem the strategy by limiting so-called stop-loss insurance, which covers unexpected, large health-care bills for self-insured companies. Lawmakers in California and Rhode Island are considering bills that would impose new rules on such coverage when offered to small employers, who otherwise would find self-insurance too risky. Some states, including New York, bar stop-loss insurers from covering small groups."

So small employers couldn't participate in stop-loss contracts? On what grounds? I guess the same grounds as those penalizing large employers from not offering health insurance.

5. Should we weaken the employer mandate?

I read the opinion as "shouldn't we MBAs have been advocating single payer?"

By the way, it occurred to me the better way to implement the "fine" would be a higher minimum wage. If an employer does not offer all employees a Federal compliant health benefit for annual $2000 or less, then its minimum wage would be $1/hour higher in 2014, rising to ~$2,50/hour higher with a $5000 health benefit annual cost. That would shift the advantage back to Papa Johns except for the small family pizza shops.

Conservatives who believe it is insurance that causes illness and accidents should like a higher minimum wage because individuals would spend the extra income buying cars and houses which will boost the economic growth. Except for those who think workers are not consumers, and consumers are not workers....

“Even though mainstream science has pretty much scrapped the notion that race has any kind of biological basis long ago, ...": see, race is just a social construct. All this stuff about drugs and ethnicity: phooey. Sickle cell anaemia? Pull the other one.

Examples like sickle-cell show only an indirect correlation with traditional race classifications. Biologically, "race" is useful only as a crude guess of a person's hereditary history when you don't have any other information. (Example: patient is black, so might have recent ancestors from Africa, so might have recent ancestors from the part of Africa where sickle-cell is common, so might have inherited sickle-cell alleles from them.) If you know pretty much anything more specific about their ancestry then the utility of "race" drops to zero. (Example: patient is black, but ancestors were from south Africa where sickle-cell is not particularly common.)

The original comment was that race had no biological meaning. What you're saying is that it's less informative than other information that might sometimes be available. That's clearly true--the information my race reveals about me is ultimately information about my genes, culture, upbringing, and experiences, and for all of those, you could get better information by looking directly at those things. But we use imperfect-but-available measures rather than harder-to-get-but-better measures for all kinds of things, all day, every day. Your doctor takes your temperature and blood pressure because those are things he can get easily, and he uses them to try to infer what's going on in harder-to-see realms, like whether or not you have an infection going on somewhere, or whether or not your kidneys are working properly. When you try to get a new job, your prospective employer looks at your resume and talks to your references and interviews you, all very indirect ways of trying to decide what kind of employee you'll be, and whether you'll be able to do the work. And so on.

#6. Sounds like a logical extension of the "Abortion Reduces Crime" debate.
Peter D

#2 no surprise here. It's the model Fox News and conservative talk radio have been riding to great success. Others have started emulating it, starting with MSNBC since 2008 and their ratings have been going up since. The limits of an information system that ranks the media on the number or followers rather than accuracy. The downfall of democracy...

"The downfall of democracy…"

Can you point to any evidence that shows people are less informed now than they were in the past?

More accurate information is more easily available now than ever before. Indeed, this is probably the reason that cable news is changing in the way it is: TV is a really inefficient way just to find out facts when we have the internet. So, in order to stay relevant, news shows have to be more about opinion and entertainment.

This Pew report says that overall, political/news knowledge stayed pretty steady between 1989 and 2007.

I don't know whether there's similar information out there for deeper knowledge--these polls draw on pretty basic knowledge and what's actively in the news day by day.

Barro's piece is just another example of why in general it is inefficient to use an indirect policy over a direct one. Health insurance mandates for employers are just partially subsidized minimum wages. It would be better to do the subsidization directly with tax credits, just as it would be better to subsidize electricity generated by non-CO2 emitting sources than to invest in specific "green jobs."

Economics 101 vs the politics of tax aversion.

In the Economics 101 vs. "politics of _____" cage match, Econ usually loses.

The mainstream beliefs seem to be:

Marginal utility of money falls off sharply with wealth and so redistribution produces big net benefits.

Apart from ability to enjoy the beach without worrying about skin cancer all races have approximately equal genetic abilities.

Poverty holds blacks back academically, that is the motivational aspects of poverty are swamped by the advantages that wealthier white parent provide for their children in academics.

Poverty motivates blacks athletically, that is the motivational aspects of poverty swamp the advantages that wealthier white parent provide for their children in athletics.

We do not need to worry about attempting to mess with people's motivations because academic motivation is not the problem access to tutors and good schools etc. is the only problem.

Comments for this post are closed