The Black/White SAT Gap

The academic achievement gap between black and white students in the United States comes up regularly in the press and in political debates. It is a deeply rooted and intractable thing, and I have long since given up on thinking that it is within our meager human powers to do anything substantial and efficacious about it. However, I also believe that it is important to have a firm grasp of the phenomenon, if only to understand what the noise is about and justify the only appropriate response: A summary dismissal of the latest wasteful education fad that will come along and be touted as The Remedy That Will Finally Work, Dammit.

This chart is the best way I can think of to put the black/white achievement gap in perspective, focusing on SAT scores.

Black-White SAT Score Gap

To put some of this in words: SAT scores rise as family income rises, across all races. However, children of the richest black families ($200,000+ annual income) have SAT scores that are, on average, virtually equal to those of children of the poorest white families (sub-$20,000 annual income). Those same sub-$20k white children outscore, by 35 points, children of black families in the second-highest income bracket ($160k-$200k), and they outscore children of comparable poor black families (sub-$20k) by 180 points. For each income bracket, the black/white gap is around 150-180 points, or nearly one standard deviation.* In sum, the lowest white group pretty much picks up where the highest black group leaves off, and achievement diverges from there.  Source: JBHE.

To put it in starker terms, the average child of a black family with demonstrably everything — a $200k+ annual income can buy you every possible educational advantage — pulls barely even with the average white child who has virtually nothing — a <$20k annual income doesn’t go very far even in West Virginia. There must be a whole lot of magic pixie dust in that poor white kid’s invisible knapsack to outweigh the after-school activities, European vacations and private SAT tutors that the rich suburban black kid can afford — but the white trailer park kid can only dream about!

An interpretive point to bear in mind is that each data point graphed above is the mean for that income and racial bracket; meaning, the actual results for all the members of each bracket probably form some kind of mini bell-shaped distribution around the plotted point, which distribution overlaps with the distributions for other income and racial brackets. There will be a few black kids from $80k families scoring 1,200+; but there will be proportionately more white kids from from $80k families scoring that high, and even proportionately more white kids from $160k families scoring that high. One should also bear in mind that the median black family income is substantially lower than the median white family income, so the gap between the median black and median white student is somewhat larger than the gap between black and white students of equivalent income brackets. However, various selection biases in test participation, as well as some probable differences in the efficacy of teaching and the prevalence of SAT-prep among otherwise-similar black and white groups I expect also confound the exact underlying difference. So I’d be perfectly content in rounding the actual SAT performance gap back to the usual 1 SD gap in underlying academic ability and achievement commonly seen between blacks and whites in the U.S.

But this next point is the crucial one for me: Over half a century of pro-black interventions — Brown v. Board, affirmative action, Head-Start, No Child Left Behind, Arne Duncan, charter schools, Teach for America, you name it — have not fundamentally closed this gap. Meanwhile, waves of poor and oppressed immigrants have historically washed up on the shores of America and, within a generation or two, have each managed to integrate educationally and socially with the mainstream, and without need for heroic federal interventions — the Irish, Jews, Italians, East Asians (anyone remember the Chinese Exclusion Act of 1882 or the Japanese internment camps of WWII?), and so forth.

It was plausible to argue in the 1950s and 1960s that black Americans simply had to throw off the shackles of former oppression (which, despite a century gone since the abolition of slavery, still plausibly existed) and that, within a generation or two, black Americans would flourish and largely merge with the mainstream, just like the Irish and the Jews and the Italians had done before. This hypothesis was substantially wrong then, as the more perceptive observers argued at the time, but it was at least plausible and certainly quite palatable to the respectable majority. So it became the law of the land and the basis for a social experiment that has gone ahead at full bore for over sixty years. Glancing at the data generated by this experiment, it is plausible that the “continuing shackles of former oppression” did have some small holding-back effect on black students into the mid-20th century, and that within a generation or two of those being thrown off black students made up that portion of ground attributable those shackles. But if that happened at all, things stopped moving within the usual generation or two, as NAEP time-series data shows:

This social experiment continues apace despite virtual lack of progress over at least the last 30 years. The record is littered with extravagant examples of abject failure and the constant drumbeat of academic testing revealing the persistence of the 1 SD gap. Despite all this, there is barely any suggestion that we should finally call time on the experiment and draw the unfortunately conclusions that nobody wants to draw: that we’ve fixed what we can, but there will be no merger with the mainstream. If trying something over and over again in hopes of getting a better result next time isn’t the definition of insanity, it is at the very least the definition of a primitive, child-sacrificing religious cult.

The most parsimonious explanation of this persistent achievement gap is, of course, the most flatly prohibited in today’s discourse… so I guess we, as a society, will continue to spin our wheels and scratch our heads and wonder why the next genius idea in closing the achievement gap (inner-city Baby Mozart? free IUDs for 12 year olds?) will barely make a whit of difference.

Others will waste their time. I, for one, just look at my first little graph and recollect the last 60 years of mostly failed experiments, reminding myself that the secret to good schools is good students — not the other way around. I conclude that there is an irreducible gap in average intelligence between black and white students, which exists on a plane that is far above our poor power to add or detract. So I depart firmly convinced that one might as well save one’s money and vote against the latest education reform fad or school tax increase… and that the rest, as they say, is commentary.

Update: Education Realist has a good insider-baseball write-up on the up-to-date status of this discourse. I highly recommend his blog.

Further Update: Due to the tendency for inconvenient truths to get memory-holed on the internet, I’ve taken a screengrab of the JBHE website and post it below for the sake of posterity.

*Second Further Update: I originally had my statistics wrong on this point. While each 800-point SAT section has a standard deviation of around 100 (or a little higher these days), the composite 1600-point SAT score (math + verbal) actually tends to have a standard deviation in the 180s or 190s. (It’s not quite the sum of the standard deviations of the two sections, given that performance tends to correlate rather strongly between the two sections — pesky old “g” strikes again!)

JBH 01-22-2009 - screengrab (sm)


One thought on “The Black/White SAT Gap”

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s