Thursday, June 07, 2007

On Ohio State's Zagorsky paper regarding wealth and IQ

Mensarefugee, maintainer of the blog DifferingPointsOfView, dug up the pdf of the Ohio State University study regarding the relationship between wealth and IQ that made a little media splash a couple of weeks ago.

The press release made a befuddling claim: IQ and income are related, but IQ and wealth are not. That left me scratching my head, because it means that those who are intelligent and make lots of money spend it more imprudently than the less endowed who don't make as much. It runs contrary to the enormous wealth gap between whites (wealthiest and highest average IQ), Hispanics (in the middle in both respects), and blacks (poorest and lowest average IQ), the advertising during wealth management shows, the relationship between education and wealth, what I read in the WSJ on each day, common sense, etc. When something so PC sounds too unbelievable to be true, it probably is.

However, the study's author, Zargosky, mercifully did use retirement savings as part of the formula for determining net worth. The problem lies in the fact that he used just about every other variable under the sun as controls. There's a clear height differential between the average NBA player and the average Chinese peasant farmer, until you take into account shoe size, pant length, femur weight, BMI, vertical jumping ability, calorie intake, and time spent playing basketball into account!

An unskeptical Yahoo news story on the study opens thus:

Pop quiz! What factors help determine how wealthy you'll become in life? I initially guessed that education, intelligence, skills, and socioeconomic origin played a role.

Ohio State economics professor Jay Zagorsky suggests different factors: "Staying married, not getting divorced, thinking about savings." Wow.
To spotlight the behaviors in the second paragraph that are clearly related to the more innate attributes of the first paragraph isn't that bothersome. Sure, it's saying "You can be anything you want to be if you work hard at it." That's not true, but working hard is a beneficial thing for most everybody, nonetheless.

Zagorsky writes in conclusion (p12):

Individuals with low intelligence should not believe they are handicapped in achieving financial success, nor should high intelligence people believe they have an advantage.

Of course if they did so they'll generally be correct, unless several factors related to intelligence are introduced as controls to attenuate its relationship with wealth to nothing (Zagorsky includes 33, every single one of which is either known to be or is plausibly related to intelligence to some degree). Yet who thinks "I'm going to climb up the ladder faster than him! But actually I'm going to be moving up more slowly, after giving consideration to the fact that his parents divorced when he was six"?

Zagorsky offers data tables based on median values by IQ range in five point increments, as releasing his entire dataset of 7,403 individuals isn't practical for an online pdf. Using these incremental averages and without controlling for any other factors, the correlation between IQ and wealth and the correlation between IQ and income are both more than .98, with income and intelligence slightly more rigorous. It is not surprising that the relationship is stronger between intelligence and income, since income is more 'purely' merit-based than wealth is (due to inheritance, astute wealth managers, etc). In light of this data, the argument that IQ and wealth are unrelated is absurd.

Those nearly perfect correlations are arrived at by treating group averages and individual samples. By individual, Zagorsky reports .16 for wealth and .30 for income, presumably without any adjustments (although I didn't see it spelled out explicitly thus).

This phenomenon is almost ubiquitous in the social sciences, and focusing on each individual data point makes it easy to portray the relationship of any two variables as meager or nonexistent.

The strongest correlation with IQ Zagorsky finds, from the 33 variables he uses, is between IQ and whether or not the participant is black. And even this inverse correlation is a modest -.35. Yet the tenacious IQ gap between blacks and whites has remained at around 15 points for at least several decades now.

What the method of presentation does do is remind us of how single-attribute statistical averages often tell us little about the individual in question, even while at the group level an overwhelming pattern emerges. If I grab the first black guy and the first white guy I see in the mall, and ask them to race, without taking anything else into consideration (height, weight, body fat percentage, age, etc), I'd be taking an almost 50/50 chance in putting money on the black guy as being the faster 40m runner. But if I take the first 1,000 black guys and first 1,000 white guys I come across, and have them pull off a massive 40,000m relay race, I'd be nuts not to bet on the black guys. So it goes with IQ and wealth and IQ and income.

Getting back to Zagorsky's conclusion, on a personal level I'm not well served--nor are readers--by assuming that we should skate through life since we have a little more upstairs than the average joe does.

Though in regards to our relationships with others, Zagorsky statement isn't optimal. It's unreasonable to assume that those who are less endowed can just as easily pick themselves up by their bootstraps as the gifted can. With advantages should come a sense of responsibility greater than that of the 'common' man.

Nor is it particularly helpful in directing us toward the more ecumenical (and controversial) goal of raising IQ scores. Doing so will attenuate wealth disparities, reduce a host of social pathologies, accelerate the rate of human advancement, and strengthen the economy.

The challenge is in getting people to embrace the desirability of a higher IQ society without casting aspersions on those who are somewhere on the left side of bell curve. I like the voluntary nature of programs like Project Prevention and child tax credits that become progressively larger as income rises rather than as it falls, and 'Lemarckian' tactics like encouraging breast-feeding and making nutritional supplements easily available to the world's impoverished. Erstwhile, the growth of entitlement societies in Western Europe and the US that've created dysgenic social pressures and perverse 'incentives' have been disastrously counterproductive.

It's a quandry I struggle with in trying to arrange a philosophical outlook: While I increasingly come to believe in the power of determinism (biological and otherwise), I see little advantage in living with a fatalistic purview. Philosophical libertarianism--basically the belief in free-will--is a more feasible way to operate on a day-to-day basis at the personal level.

But in setting public policy, determinism has to inform the debate. Holding free-will as the sole factor, to be channeled by propositionalism, is what led to the disastrous amnesty of 1986 and the fallacious belief that Amerindians from Mexico are indistinguishable from people of Northwestern European descent living in the US, save for differences in pigmentation and superficial social culture. It has also carried the US into a hopeless campaign to install a liberal democracy in the heart of the Middle East.

Why the contradiction in operating on the free-will principle day-to-day while advocating a public policy built on empirical knowledge of human biodiversity?

For one, public policy, due to its scope, is better situated to make decisions based on averages. The larger the sample, the more reliably statistical trends can be projected. The US taking in 60,000 Bhutanese refugees is guaranteed to lower its standard-of-living and IQ--a firm hiring a Bhutanese man, however, hardly guarantees that the firm will suffer for having done so.

That half of Iraqi men are married to a second cousin or closer deems a functioning nationalism virtually impossible in Iraq. In light of this, an attempt to create a nationalistic Iraq is hopeless. If I'm interviewing for a new account manager, and an Iraqi applicant appears the most qualified and competent for the spot, it makes less sense for me to reject him because I fear he'll crookedly use the assets he's in charge of to enrich his family than it does for the US to forgo involvement in Iraq due to consanguinety.

Also, consider a sports analogy. We're playing ball, and fifteen people are hanging around. We get to our last pick, and I'm deadset against taking the gangly goon who can't run or shoot well, but he and the captain are pals, so we get him. Before we start playing, I'm going to try to work with him as much as possible, and when we're running the court, I'm going to be encouraging and interacting with him.

At the national level, we should construct an immigration policy that improves the lives of natives as much as possible. When we inherit a 'liability' (the offspring of citizens), we might as well try to make him as productive and happy as he can be.

1 comment:

Anonymous said...

Do you have a link to Zagorsky's dataset of 7,403 individuals?

I did some Googling and couldn't find it.