Arnold Kling  

Charles Murray Watch

Headsets... How Real Wage Increases Have B...

Helen Ladd writes,

when first measured in the early 1940s, the gap in reading achievement between children from high and low income families was about 0.60 standard deviations. It subsequently more than doubled to 1.25 standard deviations by 2000.

These income-based achievement gaps are large. To put them in perspective, consider the black-white test score gap as measured by the National Assessment of Education Progress (NAEP) for 13 year olds...That gap was about 1 standard deviation in the 1970s, then fell to about 0.50 during the 1980s where it has remained relatively constant. As a result, the achievement gap between children from high and low income families is now far larger than the gap between black and white children.

...Even in countries such as Korea, Finland and Canada that are typically viewed as having high performing education systems, the patterns hold: achievement levels of the low ESCS children fall far short of those of their more advantaged counterparts.

The entire paper is worth reading. I got to it by following a trail from Jason Collins to Dan Pink. I recommend reading both of their posts.

Comments and Sharing

COMMENTS (12 to date)
Sam Schulman writes:

This is the argument the old Diane Ravitch used to make with white + black kids in NY public schools prewar vs postwar - reading achievement scores were a couple of clicks apart in the 20s and 30s. Something happened in teaching methods, and anyone in school in the 50s saw it happen: the adoption of progressive ed methods and the abandonment of strict, boring, rote lessons. More affluent households were harmed less because the parents/environment supported the kids outside school.
To the extent that this change in philosophy was partially a Belmont-driven deal, it would tend to support - certainly not weaken - Murray's argument. The adoption of these methods were slower in Catholic schools, and if that data were available, it would be interesting.
Anecdotal evidence: in 1956, when the worst generation entered 1st grade, the children in my class who showed problems with reading by the end of the year disappeared from my school, progressive education's fountainhead (the U of C Lab School, founded by John Dewer hisself) and were sent to the local public school for a year or two. Once the abcs had been drilled into them, they slipped back into Lab School where they could resume the pleasure of "learning how to learn" under our gifted,imaginative and eccentric teachers. But it only worked by my schoolmates' having submitted to the soul-killing teaching to the test on the part of the unrelenting and unforgiving Irish women who ruled Chicago's public school classrooms.

Floccina writes:

It is interesting that at east in theory poor people should be more motivated to work hard due to marginal utility of the results of an additional dollar (this is often pointed out in regard to progressive taxation). It seems to work that way across time (Americans used to work more that we do now) and countries (people in poorer countries often work longer hours). Given that, the author must be making the case that motivational factors of wealth and poverty are overwhelmed in education by the advantages that wealth brings.
I think we should try to objectively discern the motivational verses advantage effects in a given population also examining the genetic component all together.
To me it seems that to some people having all their needs met or having sufficient wealth, allows them to relax and choose more leisure and less work but I have no idea how many people decide enough is enough at what level. I have heard stories of children of very wealthy people how just don't work and blow the money on cars that they smash into trees and on drugs and high living. They seem to be the minority among the children of rich but they do exists so there seems to some demotivational effects due to wealth but it is ignored in the context of education.
Interesting cases to look at might be the children of big ($2 million and up) lottery winners, the children of high earning athletes and the children of rock stars and very high earning entertainers.

John Thacker writes:


The problem is that effective marginal tax rates are very high on the poor because of a loss of welfare and other government support and subsidies as they start to work. The marginal utility of an additional dollar may be high, but working doesn't necessarily give many marginal dollars. (PPACA/Obamacare, like other programs, makes this slightly worse with more phased out subsidies.)

DWAnderson writes:

I just skimmed the paper, so maybe I missed it, but what reason is there to believe that the change in achievement is caused by the schools or income levels themselves?

Why couldn't it be caused by IQ differences between low and high inocme groups becoming greater? Or a divergence in the prevalence of two-parent families between the groups?

Brian Clendinen writes:

What is the gap between Wedlock births, Divorced Families, Widowed, and two parent families. I bet there would be very little achievement gap on income or race if wedlock births where taken into consideration (achievement gap can almost totally be explained by wedlock). It all has to do with the condition of the family the child is in. Race and Income are a fairly irrelevant achievement gap measurement. The only achievement gap metrics that really matters from a strategic policy standpoint is parental status, sex and type of school.

Jeff writes:

Mating is more assortive than it used to be, and IQ is heritable, so this is what should be expected.

Josh S writes:

What I get out of this is that government monopolies have little incentive to serve poor communities particularly well. Decades of service this bad would have bankrupted any business.

Lee Kelly writes:

Conscientiousness and intelligence are great predictors of income and wealth. Both are highly heritable. The gene pool is being divided.

kczat writes:

I look forward to reading Bryan Caplan's response to this. It seems that twin studies would have a lot to say about the extent to which this is genes versus environment.

Elvin writes:

Wouldn't assortive mating create a wider distribution of outcomes, not a bipolar one? Even with a slight regression to the mean effect, smart people mating just means that a few more geniuses occur than if mating was random; at the other end of the curve, it would mean a few more some really dumb people result. With a widening Bell curve, the range of outcomes between the top and bottom of the curve should increase as Ladd asserts. Other factors as the commentators have mentioned--quality of school and marginal tax incentives--are possible.

Still, in between Fishtown and Belmont is the bulk of the Bell curve with a a lot of middle class people. Under assortive mating, these middle class people are producing middle class progeny, but maybe with a slightly lower standard deviation.

I just don't see how a bipolar distribution results. A wider distribution, yes; but not a bipolar one.

Shayne Cook writes:

To Floccina:

Some very good research work has already been done in exactly the areas you refer to in your comment - see "The Millionaire Next Door", by Stanley and Danko. It's very enlightening.

Ghost of Christmas Past writes:
consider the black-white test score gap as measured by the National Assessment of Education Progress (NAEP) for 13 year olds...That gap was about 1 standard deviation in the 1970s, then fell to about 0.50 during the 1980s where it has remained relatively constant

But the SAT score gap has not closed at all. Possible explanation: political emphasis on "gap closing" has lead primary schools to push blacks harder and slack off on whites. This affects intelligence much less than it affects memorization of stuff tested by NAEP.

Comments for this entry have been closed
Return to top