I don't have any stats handy, but I've read several times that after the civil war, blacks saw steady improvement in their economic status, on similar (albeit slower) trend to various immigrant ethnicities who faced initial integration troubles, but that this progress stopped and then reversed in the '70s. This reversal correlated very strongly to the rise of out of wedlock births among blacks. I've read that, if you correct for single-parent/dual-parent families, the difference in school performance between all ethnic groups almost completely disappears, even if you don't correct for things like income level or the education level of parents (although those two things also correlate with single/dual-parent status, so correcting for it tends to correct for them to some degree). Did slavery cause marriage and stable families to decline among blacks over a century after the 13th amendment? It seems to me that, just as our society was finally putting "the legacy of slavery" to rest, something else happened to hurt the black community even worse. If we can figure out what it was, we'll be well on our way to closing the gap. Unfortunately, so much damage has been done at this point that I don't think that simply removing the original cause will fix the problem--extra measures will be needed to repair the damage.
You talk about "norms" but as was pointed out in another HN thread (http://news.ycombinator.com/item?id=2031655) drug use is essentially a "norm" in the USA. The problem is that only certain segments of the population get punished for it, which leads to not finishing school, not being able to get a job, etc. It certainly is a cause.
I wasn't around at the time, but I think that the social movements of the '60s and '70s threw the baby out with the bathwater. There were definite wrongs that needed to be fixed (segregation being the prime example), but in the process of fixing them we inadvertently damaged key aspects of our societal structure. I think that this happened because of the way our culture lumps together completely unrelated positions into strongly correlated groupings that we label "conservative" and "liberal." As a modern example, consider the fact that people's positions on gun control, abortion, and taxes all tend to correlate very strongly in our country, even though all three are orthogonal (libertarians being an example of people who buck this trend to some degree). In the '60s and '70s, bad things like racism got lumped in with many of the good things we collectively label as "family values," to the point where many of the people fighting against the bad stuff saw the good parts as being part and parcel.
This carries over today: my most liberal friends consider "family values" to be an ugly euphemism for "racism," "homophobia," and "misogyny." They don't make any distinction for the good parts like "get married before you have kids," and "marriage is a serious and important commitment which requires a lifetime of compromise and sacrifice." They think that the bad parts are inherent aspects of the culture, which implies that the only way to get rid of them is to radically change the entire culture. I've even read positions more extreme than those of my friends: people who think the parts I label good, like the institution of marriage, to be part of the problem.
I'm Irish, and I'd question the accuracy of that statement - whether the US was unique in that regard.
You could argue over the specifics of what it means 'to try stamp out literacy', but for 100s of years here (Ireland), the native language (Irish/Gaelige), and education through it, got a very tough time from the occupying British government.
More generally, the country had a tough time. The majority of the population (Catholics) had little to no opportunity, and little education ('the penal laws' etc). Many of those who could left as emigrants (going to places such as USA).
Now, Republic of Ireland - the OECD nation - has been around less than 100 years, and we've made big strides in this time. Many peoples parents are now college educated; but this is probably the first generation where that's true.
So, what does it mean to compare the population of Ireland 'corrected for the demography', as is done on the last chart in that page? And then make statements about the legacy of slavery in the USA, using that comparision?
I really think the analysis lies in tatters at that point.
While this post is - clearly - raising interest in the measurement of education (a good thing) the method used is just much too crude to possibly draw the kind of conclusions the author, and others, are making.
There's just too much historical differences between what it means to be an immigrant in different countries, and too much different history, for such crude analysis to support the conclusions drawn.
"Among those 34 OECD nations, only the United States spent centuries aggressively trying to stamp out literacy among a major part of its population. The legacy of that benighted history lives with us today, although our “reformers” work very hard to avoid such painful discussions."