« Assorted | Main | 10 Things You Shouldn't Buy New »
The best sentence I read today
On average, taller people earn more because they are smarter.
Maybe it's not greater self-esteem.
Posted by Tyler Cowen on August 22, 2006 at 09:21 PM in Economics | Permalink
Comments
Why don't they just look at adult IQ scores and height?
Posted by: joan at Aug 22, 2006 10:03:22 PM
That doesn't actually surprise me, given that stunting - defined as children under age five who are under height for age - is a measure of chronic malnourishment. From the Center for Global Development's new report, Measuring Commitment to Health (final version available soon):
Improving nutrition also has a long-term effect on household poverty, as properly nourished children are likely to earn more when they reach adulthood. This mainly is a result of the education effects of improved nutrition, which improves developmental levels in infancy, leads to more and earlier school enrollment, improves cognitive function, and can decrease absenteeism due to illness. A 10 percent increase in stunting in the average Ghanaian child correlates with a 3.5 percent increase in the age of school enrollment, which a small improvement in height-for-age in Pakistani children is correlated with two percent increase in school enrollment rates for boys and a 10 percent increase for girls. Ultimately, this improved education can significantly increase earning potential and bring people out of poverty.In addition to delaying school entry, stunting also can hinder cognitive development, which further hurts education prospects. A 1999 study found that, in Filipino children, both moderate and severe stunting in the first two years of life had a significant negative effect on cognitive test scores in late childhood even when researchers controlled for the amount of schooling received. Malnutrition also can significantly hinder motor skills development in young children, which may limit their ability to perform physically demanding tasks in the future.
Granted, our report was focused on developing country contexts, but the general premise holds.
Posted by: Jessica at Aug 22, 2006 10:13:03 PM
With all due respect, of course the general premise does not hold. The difference between "tall" and "short" children in the United States (at least among people who grow up to take the kind of jobs talked about in the NBER paper) is not the result of "chronic malnutrition." Therefore, any conclusion based on studies from the developing world are irrelevant to understanding the connection between height and intelligence (if it exists) in the developed world.
Posted by: William Goodwin at Aug 22, 2006 11:04:25 PM
I think there's still plenty of room for a correlation between nutrition, height, and intelligence in the US.
Posted by: Stephen Duncan Jr at Aug 22, 2006 11:19:33 PM
I'm not an expert in this area but I believe that there is a gap in heights between rich and poor adult Americans of about 1 inch. This wouldn't be interesting by itself except that the gap used to be much greater (in modern India, the correlation between socioeconomic status and height is glaringly obvious). So this suggests that nutrition for poor Americans has improved relative to rich Americans significantly but is still not the same. In developing countries, poor people still have a large nutritional deficit that causes stunting. There is also a fairly strong, though certainly not perfect, correlation between height and IQ in the U.S. Genetics could be at work here as IQ almost certainly has a genetic component and height likely does as well but a rise in average heights in a country tends to correlate with a rise in average IQ scores, suggesting that nutrition is the link between these two.
Another somewhat weaker piece of evidence for the link between height and childhood nutrition is that average heights are higher in northern Europe -- the Netherlands holds the world record -- than in the U.S. and it is also true that babies tend to be healthier in northern European countries than in the U.S.
Posted by: Mark at Aug 23, 2006 12:05:34 AM
Mark -- even if nutrition is the same for everybody, that last remaining inch could be explained by some combination of the old "tall people earn more due to social reasons" story combined with the new "genetically tall people are smarter and so earn more money" story. I might add that I have a hard time believing the new story.
Posted by: Alex F at Aug 23, 2006 1:57:57 AM
Stunting due to poor health is very probably the reason for lower intelligence in the population of shorter people.
In a nutshell, smaller people contain healthy people who are genetically small, but also genetically tall people who have been stunted by stresses such as poorer nutrition (including nutrition in the womb) and serious childhood infections (which produce anorexia and weight loss). (A few tall people are unhelathy due to growth hormone excess, but this is very rare.)
This is likely why there is a strong social class gradient in height (higher classes being taller), and why increased height also correlates with better health. Indeed, this is probably why tallness is attractive. This attraction has probably evolved as a way of selecting mates of higher reproductive value (on average).
So, by being attracted to taller people, humans are also attracted to (on average) healthier people. Healthier people probably are less likely to be carrying deleterious genetic mutations, and are less likrly to have their reproductive and child-caring lifespans shortened: so they will survive long enough to have more children and raise them to maturity.
Posted by: Bruce Charlton at Aug 23, 2006 3:24:27 AM
I am 6' 1" and 143 IQ, Stick that in your data points.
"Maxthewise" is from an inside family joke, OK?
Posted by: Max at Aug 23, 2006 4:44:20 AM
There are, of course, notable exceptions. I realize the quote says "on average", but the overall, average intellectual and financial success of Ashkenazi Jews, many of whom have tended to be shorter than average, is remarkable. For more, see the comments at
http://www.eclectecon.com/posts/1156297918.shtml
Posted by: EclectEcon at Aug 23, 2006 5:14:32 AM
The difference isn't necessarily going to be directly due to nutrition - a child in a family that buys cheap, processed food that isn't an optimum food for growing infants is also a child that is less likely to have parents with a college education, less likely to have books around the house, less likely to have the expectations of achievement.
Posted by: Dr Maybe at Aug 23, 2006 6:56:05 AM
William - malnutrition probably isn't the only factor, but it could be a contributing one. Sorry if I seemed to overstate my case.
Posted by: Jessica at Aug 23, 2006 7:05:37 AM
eclectecon: There are, of course, notable exceptions. I realize the quote says "on average", but the overall, average intellectual and financial success of Ashkenazi Jews, many of whom have tended to be shorter than average, is remarkable.
My reply: Any scientific theory implicitly include 'ceteris paribus' - in this instance, the control would be to study the correlation of height with intelligence _among_ Ashkenazi Jews.
Posted by: Bruce Charlton at Aug 23, 2006 8:16:50 AM
Bruce, that doesn't work in this case. You can't make a broad statement about the overall correlation of height and intelligence, and then say, "But that's controlling for other factors that might make our conclusion false." If -- for the sake of argument -- Ashkenazi Jews are both shorter, on average, than the rest of the population and smarter, on average, than the rest of the population, then that suggests that the superior smartness of tall people may, in fact, not be real. You can't, in other words, control for height if height is one of the variables you're trying to measure.
Posted by: William Goodwin at Aug 23, 2006 11:16:51 AM
Replying to William Goodwin - Sorry, I don't follow your point. I'm just proposing standard scientific methodology. Controlling for a particular Jewish ethnicity isn't the same as controlling for height - I guess it controls for some genetic factors and some cultural (eg dietary and educational) factors.
Actually, I think I can frame the hypothesis more precisely: that taller height and greater intelligence are outcomes of better health (including good genes) (assuming that relevant confounders are controlled).
The reason is that intelligence (brain development etc) is reduced by poor health/ deleterious genetic mutations/ and poorer nutrition (especially in the womb and childhood); and so is height.
Posted by: Bruce Charlton at Aug 23, 2006 12:05:49 PM
Bruce, if you're arguing on the one hand that height is the product of good genes and good nutrition, and then you want to control for genetic/nutritional factors (that is, those that are specific to Ashkenazi Jews, assuming they exist), then you are, in fact, controlling for those things you're supposedly trying to measure.
To put it differently, you're saying that tall people are smarter because the same factors that determine height also determine intelligence. If we look at a particular group of people and find that they are both shorter than average and smarter than average, then that is an important piece of evidence against your thesis. And you can't do away with the counterevidence by saying, "Well, let's control for all the things that these short, smart people have in common and then see if the thesis still holds, because the thesis is a statement about humans in general, not a statemtn about humans within particular ethnic groups.
Posted by: William Goodwin at Aug 23, 2006 12:21:22 PM
William. I am not engaging in any sleight of hand here! This is just how people do epidemiology.
Let me try another example. Men are taller on average than women. This is true, biologically. Humans as a species are sexually dimorphic - like all the other great apes - but not with such a large differential.
Of course, there are some societies in which the average height of women is taller than the average height of men in other societies - for example the average Norwegian women is (I am pretty sure) tall than the average pygmy man. That is because height is influenced by several factors as well as sex.
BUT - if you were to control for all relevant, height-infuencing factors, then _all_ men would be taller than _all_ women.
So, controlling for some of the height-affecting factors that make Ashkenazi Jews smaller than (say) Norwegians, you should still expect to see the relationship between height and intelligence.
Posted by: Bruce Charlton at Aug 23, 2006 1:17:11 PM
"So, you're saying the humans are dumb, yet... tall. How is that even possible? How can anything tall be dumb?"
-- Almighty Tallest Red ("Invader Zim")
Posted by: Xellos at Aug 23, 2006 1:29:41 PM
Bruce, you are engaging in a sleight of hand. You're assuming that the "height-affecting factors" that make Ashkenazi Jews smaller than Norwegians (assuming they are), have nothing to do with the fact that Ashekenazi Jews are also smarter than Norwegians (again, assuming they are). But you've already argued that the genetic/nutritional factors that make people taller also make them smarter. So how can you control for the former without also controlling (inadvertently) for the latter? Again, if you have (for the sake of argument) a large group of people who are both smaller and smarter than other humans, why would we assume that tallness and intelligence were correlated?
Posted by: William Goodwin at Aug 23, 2006 2:03:41 PM
When I was young I was smart enough to know I had to grow tall to be successful, so I did.
Posted by: JOhn Fembup at Aug 23, 2006 2:28:11 PM
hmm, where do i fit? i was always short as a child, like maybe one kid was shorter than me in school and such... then i hit a big growth spurt around 15 and now im like sorta above average at 5'11... so does this mean that im a dumb guy in a smart mans body? or did i become smarter when i grew? or? or? is this the lamest study on earth?
also whats the deal with poor nutrition being blamed for shortness? maybe in developing countries? but i seriously doubt my nutrition jumped by leaps and bounds when i was 15. is the assumption i finally had a salad, drank milk, and fell in love with a multi-vitamin?!
Posted by: shortThenTall at Aug 23, 2006 5:25:56 PM
I couldn't read the paper, but I hope they controlled for age. I remember the first couple of sentences in my first statistics course: "what if we link bigger feet to higher reading ability in children...big footed children are smarter?...no, children with bigger feet tend to be older and have higher reading levels because they have learned how to read better since they've had more time on this planet."
I'm sure they controlled for age, but it would be comical if they didn't.
Posted by: Scott W at Aug 24, 2006 12:59:32 AM
"processed food" is not necessarily less nutritious than "organic" or other so-called natural food. In fact some processing is necessary to prevent spoiling and infestation, without which the self-styled pure natural food is actually much more dangerous than properly processed food. I'd rather eat Cheez Whiz than a wild berry or mushroom with unknown amounts of toxins in it.
Posted by: Robert Speirs at Aug 24, 2006 11:07:01 AM
BUT - if you were to control for all relevant, height-infuencing factors, then _all_ men would be taller than _all_ women.
Not true. One of the relevant factors is sex, so if you control for all relevant factors you will also control for sex, and men would be the same height as women.
Bruce, William Goodwin is right.
Posted by: Ben Tillman at Aug 24, 2006 9:40:10 PM
Reply to Goodwin and Tillman.
I may not be making myself sufficiently clear. I am discussing the biological theory that height and intelligence correlate because they are both related to health - specifically that poorer health reduces both height and intelligence. One epidemiological way to investigate this is to look at populations, controlling for various other aspects that influence height - for example one might study a mixed national population in several coutries, and several more homogeneous ethmic populations such as Ashkenazi Jews. If in all these studies height and intelligence are still correlated with health, then the theory may stand, and further tests be devised.
There are lots of other ways to test the theory - and the more detailed the theory was spelled out (eg at the genetic, hormonal, neuroscientific levels), the more rigorous the tests could be. At some point the theory would become more specific, or would break up into sub-theories - for example some kind of disease might affect height but not intelligence or vice versa. It might even be discovered that some speciifc factors seem to have an excepionally large influence - for example serious childhood disease seems to be be a particularly potent cause of reduced height in adults.
Maybe William Goodwin's point is that if there were enough Ashkenazi Jews in a population, then this would obliterate or reverse the statistical association between height and intelligence in that specific population? That is true. Maybe his point is that there are many factors which influence height and intelligence. That is also true.
But neither of these facts refute the hypothesis that height and intelligence are both influenced by health - which happens to be the hypothesis I am interested in! Nor would it be invalid or mistaken to test this hypothesis in sub-groups such as Ashkenazi Jews.
Posted by: Bruce Charlton at Aug 25, 2006 3:03:41 AM
PS: to the above - it strikes me, somewhat belatedly! that William Goodwin is probably interested mostly in the hypothesis that Ashkenazi Jews are more intelligent than other groups, and that (if true) this may be caused by whatever (gene?) it is that makes them shorter. An interesting idea in its own right - and of course quite different from my hypothesis about health - however it may explain the confusion...
Posted by: Bruce Charlton at Aug 25, 2006 3:45:19 AM
