Posts from — August 2009

Greg Mankiw is still defending his SAT scores post while completely missing the point. This time he shares a graph that shows children’s IQs being correlated with their biological father’s income, for both adopted and non-adopted children. The suggestion being that test scores are largely inherited and high income is just an effect of raw intelligence. However, he ignores the fact that the IQ and SAT tests are not the same! They are trying to measure different things, IQ attempts to measure some sort of innate intelligence, while SAT tries to measure college preparedness with regard to things like literacy and mathematical knowlege. Certainly there is an interaction between the two, but the effects of preparation and education should be reflected in the SAT to a much greater degree.

His post title “Test Scores and Biological Father’s Income” seems to try to blur the issue and he never mentions differences between the tests, as far as I can tell. To be clear, IQ was certainly a factor in the SAT scores, but his claim that it’s “the key omitted variable” isn’t supported by this evidence.

Greg Mankiw, whose post about test scores and household income elicited my response, seems to be bewildered about why people are pushing back against his post. It’s really too bad that he doesn’t have comments or a public email on the blog, as the reason is quite simple: yes, genetics and innate inherited intelligence have some effect on IQ, and IQ does correlate with income to some degree, but the claim it the key omitted variable is making an even stronger statement. It’s saying that these genetic effects outweigh the affordances that affluent people have in society. It’s making a social darwinist claim that society is roughly able to sort people by intelligence.

Stuff like this make me wish I had found better conservative bloggers…

Eternal Earth-Bound Pets is an interesting variation of the “rapture insurance” idea that I’ve heard a few times. It works like this:

[...] When the Rapture comes what’s to become of your loving pets who are left behind? Eternal Earth-Bound Pets takes that burden off your mind. [...] We are a group of dedicated animal lovers, and atheists. For $110.00 we will guarantee that should the Rapture occur within ten (10) years of receipt of payment, one pet per residence will be saved. Each additional pet at your residence will be saved for an additional $15.00 fee. A small price to pay for your peace of mind and the health and safety of your four legged friends.

The expected value of this insurance is definitely quite different from the perspective of the receivers and providers in this case. I’m not sure that the atheists would be capable of pet-sitting in the case of rapture, so I don’t know that this would work.

(via BoingBoing)

Greg Mankiw called this graph “the least surprising correlation of all time”:

(SAT scores by family income)

I agree that it’s unsurprising, but for different reasons. Certainly, this correlation shouldn’t imply direct causation, as Mankiw states:

Suppose we were to graph average SAT scores by the number of bathrooms a student has in his or her family home. That curve would also likely slope upward. (After all, people with more money buy larger homes with more bathrooms.) But it would be a mistake to conclude that installing an extra toilet raises yours kids’ SAT scores.

Agreed, but I don’t agree with this:

The key omitted variable here is parents’ IQ. Smart parents make more money and pass those good genes on to their offspring.

While I’ll admit that parents’ IQ would show some correlation with test scores, I don’t think that this is the “key variable”. First of all, the SAT does not measure IQ, it attempts to measure language and math skills. I’d expect there to be a better correlation with some measure of the parents’ educational achievements.

August 14th’s must-listen episode of This American Life touched on this subject in the first act. Childrens’ academic success (and success in general) tracks very well with how much their parents talk with them when they’re very young and how much they read to/with them. Parents who read on their own are more likely to do these things.

Additionally, there’s plenty of training that goes into doing well on the SAT. More affluent parents can afford to pay for tutoring for their children and, more importantly, move to a town with a better school system. If standardized test scores were just measures of innate ability, then it would be absurd to measure the performance of a school system based upon them, and you could just test a child once. (Not saying that they’re a great way to measure, but I think that there are other problems with that.)

Growing up in Avon (suburban CT), I saw plenty of very wealthy homes produce terrible students, but the student body as a whole did well largely because we had a good school system, involved parents, and a culture that said it was OK to be smart.

So, yeah, I’m in the ‘nurture’ camp on this one.

I was talking to my Mom, who teaches nursing, about health care last weekend. She made the point that a sign of the quality of the American health care system over that of other countries is that people from Canada come here for surgeries sometimes (and pay for it!). This does happen, but I’m not truly convinced by the implication of it.

One meaningful thing that I think a whole lot of people involved in the healthcare debate can agree on is this:
The American health care system is the best health care system in the world for wealthy people.

While I’m not quite sure that this is 100% correct, it’s certainly a lot more correct than it would be without the last few words.

Despite our expenditures (more on that later), we’re 50th in life expectancy, at least 33rd (lowest) in infant mortality, and the WHO ranks our overall health system as 37th in the world in overall quality.

Some more information about the WHO’s methodology is here, and in fact the US did well in one of the components of the performance index:

Responsiveness: The nations with the most responsive health systems are the United States, Switzerland, Luxembourg, Denmark, Germany, Japan, Canada, Norway, Netherlands and Sweden. The reason these are all advanced industrial nations is that a number of the elements of responsiveness depend strongly on the availability of resources. In addition, many of these countries were the first to begin addressing the responsiveness of their health systems to people’s needs.

I’m not entirely sure what this means in practice, but I can’t think of any other (good) health stats in which the US leads the list.

No one measure is a good summary of the quality of a health measure, as the systems are hopelessly complex. The WHO stopped doing the index in 2000 for this reason (and maybe some political ones..). Life expectancy seems like a good measure in general, since that’s a main point of these systems, but the inputs are completely different. Americans live differently, eat differently, drive more, and shoot each other more than most countries on the list.

The ‘best’ system must take cost into account, something the WHO did. A healthcare system that lest you live a year longer, but at twice the annual cost for your whole life probably isn’t an improvement.

The majority of the healthcare reforms being discussed are actually health insurance reforms. I think it’s important to take a step back and think about what insurance is, since there’s more confusion about that than you’d expect.

Insurance is a way to protect against an unlikely, catastrophic event. Yes, health insurance does more than just that, but this is the core of any insurance. The way this works has a lot to do with expected value. The expected value of an event is the probability of that event happening multiplied by the value (or cost) of its occurrence.

A deck of cards can provide an easy example. Let’s say we play a game where you pull a card from a deck, and if that card is an eight of clubs, you have to pay $200,000 (or some infeasible amount). It’s a crappy game, I know, but that’s life. But instead of risking the 1 in 52 chance of financial ruin, you can buy insurance. The cost of this is based upon expected value: there’s a 1.9% chance of the catastrophic event, and that event costs $200,000. Thus, insurance would cost about $3,846 (the product of those numbers) for the insurance company to break even given enough people/time.

You, the insurance customer, benefit from the insurance because you no longer have to worry about complete ruin from the game. The insurance company charges a bit more for both profit and overhead, and if their estimates of the expected value are good and they have enough customers so that the law of large numbers takes effect, they do well.

What effect does profit have? You’d think that since Insurance company A is charging [expected value + some profit/overhead] and B is charging [expected value + some other profit/overhead], eventually the profits would just drive down to zero and you’d just want to create a co-op non-profit insurance group. In other words, why wouldn’t I want to just compute expected values myself and pool money with the people at my workplace, university, or whatever? Well, I’ve clearly oversimplified the problem, because an important part of the service provided by any insurance organization (government or private) is a verification of the claims when they’re paid out. If nobody is stopping people from claiming all sorts of expenses in our insurance, the costs rise for the honest people. Insurance companies are supposed to control costs.

Different people have different probabilities of these catastrophic events, of course. So, returning to the analogy, let’s say you were given a bad deck and have a proclivity for the event for some reason and there are now 2 eights of clubs. Thus, you have a 3.8% chance of failure, and your expected value is $7,692. If you have an increased chance of pulling the bad card, the insurance company needs to charge you more. If the insurance can’t know if you have this increased risk, they need to charge everyone slightly more to account for this (even those with a regular deck).

A real preexisting condition, like someone applying for health insurance when they already have cancer, breaks this whole system, though. In the analogy, this situation would be one in which the eight of clubs was the only card in the deck. Thus, the expected value is $200,000, and the insurance company has to charge at least that much to break even. So, the entire insurance proposition now makes no sense for everyone involved. There’s no reason for you to get insurance in this case, and an insurance company wouldn’t be able to provide a worthwhile product.

Thus, the people with known conditions like this can’t really be covered by insurance, but if we as a card-game-playing society decide that nobody should have to lose the game like that, the cost could be amortized among the insurance payers over time, raising their costs. It seems like all of this applies no matter how public or private the insurance options are.

So, to review, here are the things I think about insurance right now:
- Insurance provides a valuable service to customers.
- Insurance organizations (public or private) are responsible for keeping costs under control and curbing abuse.
- Insurance organizations seeking to reduce costs or increase profits have an incentive to deny payouts that may be legitimate.
- Insurance doesn’t make sense for people who have expensive preexisting conditions.

I know that this is very simplistic, but if my understanding is fundamentally wrong, please let me know.

I’ve been wanting to write about health care for a while, and now I finally have the time to get started.

It’s tempting to just jump in with a small post about some tactical issue in the debate, or make fun of Sarah Palin’s death panels, but I think I’m going to try to avoid that and instead try to keep it more abstract. I don’t really have a strong opinion about what should actually be done, but maybe as I write one will emerge.

Since I literally didn’t know what the term “single payer” meant a year ago, there will doubtlessly be errors in my posts, so please submit corrections and criticism.

I always find those “I won’t be posting for a couple of days” blog posts to be rather narcissistic, but here I am, pretty much writing one (but for a longer period of time). I’m in the midst of a very busy summer, described below, and I likely won’t have the time to write much for a little while.

My weekends from late July to early September consist of:
Europe (conference)
Europe (vacation)
Europe (vacation)
Family visiting
Family visiting
Trip to Pittsburgh and/or moving (within SF, not to Pittsburgh)
Klump visit and maybe moving

It’s certainly been a great summer so far, but I must admit I’m looking forward to things settling down a bit, too.

If you’re looking for some reading, check out my google reader shared items in the meantime.

Don’t worry, I won’t forget to write about “death panels” when I get back. Death Panel would be a good band or movie name, by the way.