Arnold Kling  

Reverse Ability Bias

Sacerdote, Feyrer, Kids, and G... The Risk Disclosure Problem, R...

David Card writes,

estimates based on supply-side innovations tend to recover returns to education for a subset of individuals with relatively high returns to education.

Let me try to put that into English.

The conventional wisdom is that when somebody looks at the relationship between years of schooling and earnings, they over-estimate the impact of schooling because of ability bias. That is, people who have more schooling have more ability, and so you might naively attribute all of their higher earnings to more education when in fact much of it may be due to ability. A more sophisticate approach would attempt to correct for ability bias.

Card's paper, referred to in the Goldin-Katz book, summarizes some attempts to address this. But it uses Fancy Econometrics, which makes the results suspicious, and my interpretation of them even more so.

The way I think of it, suppose we have two types of people. A-plus Adrians have lots of ability. Pass-fail Pats have mediocre ability. If we can find a control group of people, where presumably mostly Adrians go to college, and compare it with an experimental group where we have reason to believe that the college attenders include a higher proportion of Pats, then we can get an idea of how much ability bias there is. The experimental group would have a lower college wage premium because of all of the Pats who managed to sneak into college.

In fact, what studies like this show, according to Card, is that the experimental group tends to show a higher earnings/education gradient than the control group. It appears, therefore, that what we have is reverse ability bias! Perhaps people with more ability get more consumption value out of school and actually derive less investment value at the margin. That is, Adrian uses an extra year of schooling to go for a Masters' in art history, while Pat takes courses in writing and computer science at the local community college.

That's the story, assuming I get the gist of the paper, and assuming you believe the results. Are the experiments convincing?

In Card's own study, you can think of the experimental group as consisting of folks who grew up near a four-year college. If that makes it more convenient to attend college, then you should see more Pats among that group attending college. If the college/high school wage premium is higher within that group, this tells you that at the margin Pat gets more out of college than Adrian.

But as I said, Card uses Fancy Econometrics. That means that, as far as I can tell, the higher returns to education in the experimental group could be coming from higher returns to more years of high school. They do not necessarily have anything to do with college at all. I really wish folks would use more descriptive statistics and fewer regressions.

Comments and Sharing

COMMENTS (9 to date)
meep writes:

...and if you lived near a 4-year college, it may be because your parents work there.

There are lots of reasons those living near a college may have a different distribution of ability than those who live farther away. And it may be in the opposite direction.

Alex J. writes:

I know nothing of fancy econometrics, but are the Adrians more likely to go into academia? This would reduce their income. During the Summers controversy, Philip Greenspun pointed out that:

Adjusted for IQ, quantitative skills, and working hours, jobs in science are the lowest paid in the United States.

Arnold Kling writes:

I didn't get into it in the post, but Card controls for family-background variables. The only question would be whether his controls are adequate. That is, the parents' education level is measured and controlled for. However, some further unobservable dedication of the parents to education or culture that might lead them to move close to a college cannot be controlled for.

Steve Sailer writes:

Does he control for work ethic?

If you have two sets of people, one with a lower average IQ, but they end up with equal amounts of schooling, the the lower IQ folks will likely average better work ethics. And that should stand them well in the workplace.

Arnold Kling writes:

As long as work ethic is randomly distributed between the people growing up close to a college and the people growing up far away, the work ethic issue will not affect his results.

Libra writes:

Does anybody know if there has been a study of the returns to college education controlled for SAT score ( or IQ)? I was doing some major Googling trying to find such a study, but to no avail.

Libra writes:

Someday I'd like to start a social science audit web site. For papers like these, the site would publish all the raw data, plus the actual code for the econometric models. Visitors could then play around with the numbers and examine all the hidden assumptions and fudge factors. They could alter and re-execute the model right in the browser. If a visitor could plausibly argue that the weightings/choice of variables could be something different, and that change results in dramatically different conclusions, then the entire model is worthless. I suspect if social science papers were subjected to such scrutiny, 99% of them would turn out to be garbage. Perhaps that's why it has not been done.

Outside of calculation biases, I can think of numerous holes in this model: 1) college towns usually have higher costs of living, which drives wages up. 2) universities have been a major growth industry ( due to credentialing laws, government subsidies, and ridiculous endowment growth) and thus areas near colleges might have had increasing wages 3) University towns suck in a lot of people of high ability, which again, drive up the wages.

Finally, even if there is a correlation, we still wouldn't know if it's due to increased productivity, or due to credentialing laws, the growth of the civil service, and the Griggs decision. We need separation of schooling and credentialing in this country!

BGC writes:

This kind of study simply must control for IQ - controlling for years of education is inadequate.

But I have been looking for comparative studies of educational interventions which control for IQ, and I haven't yet found any.

IQ is more than 100 years old, and is the best validated psychometric measure - by far. So why are there thousands of studies which dont control for IQ or even discuss the subject? What the heck is going on?

Oh - I forgot. People who work on IQ are evil.... (irony).

By deliberately neglecting IQ our society is wallowing in self-righteous ignorance.

Slam writes:

This kind of study simply must control for IQ - controlling for years of education is inadequate.

The whole point of instrumental variable estimation or looking for 'natural experiments' is to address the problem of ability bias. Besides, IQ scores for individuals are not as freely available as wage and education information. And there have been numerous papers on returns to education that have included measures of ability. No paper in any respectable econ journal nowadays would get published if the author does not pay attention to ability.

Comments for this entry have been closed
Return to top