Yesterday I happened to come across a tweet bemoaning the state of reading among fourth graders. Pretty standard stuff. But as things so often do, it got me curious. Here are reading proficiency scores for fourth graders over the past 50 years:
I'm showing this for white kids so it isn't affected by demographic changes over the years. The main problem is that to go back so far I had to look at the NAEP's long-term test, and the LTT doesn't use normal language like proficient or advanced to describe reading ability. It shows only bands of scores, and then describes those scores using incomprehensible gobbledegook. It's truly weird. I did my best to convert this into words that seemed like reasonable translations.
In any case, the thing to notice is that there's been hardly any change—and what there is suggests a slight improvement. Whether it's phonics or whole language; books or iPads; and through all the change from TV to texting to social media—nothing changed. On average, the kids turned out the same no matter what.
And maybe that makes sense. About 25% of kids have an IQ above 110. They're probably all pretty proficient readers. Likewise, about 25% of kids have an IQ below 90. They're probably all either very weak readers or just flatly illiterate. And then you have the 50% in between who end up at various levels of "OK."
And maybe that's that. Maybe the fads and the trends and the quality of teaching only matter at the margins. About a quarter of the population is always going to be barely functionally literate.
The results for Black kids are a little different. Black students score significantly lower than white kids overall, but their proficiency levels have improved somewhat over time. Not a lot, but enough to suggest that changes in teaching have some effect.
With that in mind I began to wonder which states do the best job of teaching Black children to read. Looking at raw scores isn't very helpful since states differ widely on poverty levels and we know that poverty affects school outcomes. But we can control a bit for that by looking at how close Black scores are to white scores in each state:
There's a huge range. Vermont's Black kids are only 9 points behind, which isn't too bad. In Washington DC they're an astonishing 54 points behind.
Even if you remove those two outliers the range is 24 points. That's a helluva lot, roughly equivalent to 2½ grade levels.
Nor does it shake out the way you might think. Plenty of Southern states do fairly well: Texas, Florida, Kentucky, and North Carolina. And plenty of northern states do poorly: Rhode Island, Connecticut, New Jersey, Minnesota, and Wisconsin. It's not clear what the differences are, but it's not much related to ancient patterns of slavery and racism.
"It's not clear what the differences are, but it's not much related to ancient patterns of slavery and racism."
I would try poverty and lack of family cohesion.
Speaking from experience, segregation happens organically, even in my former state of Maryland. That segregation also segregates the money into different taxing districts.
"I'm showing this for white kids so it isn't affected by demographic changes over the years. ..."
And what effect have these demographic changes had on test scores? Ignoring unfavorable demographic changes doesn't make them magically disappear.
DoD?
Dept of Defense schools are known for high scores among non-white kids. They have a standardized curriculum, so kids get a consistent education even as they move from place to place. Moving from school to school because of housing instability is one reason low income kids do so poorly in most schools. Only the DoD can really correct for that by forcing standardization. Kids in the US just end up bouncing from district to charter and back again, sometimes attending more schools than they have years of schooling. Because poverty correlates with race, that leaves a lot of Black kids with a really haphazard education.
The other factor is that the military weeds out a lot of the worst parents. You have to be able to read, follow directions, not have severe mental illness, and not be high all the time to remain active duty. Public schools in the US have kids being raised by meth addicted schizophrenics. DoD schools do not.
"Plenty of Southern states do fairly well: Texas, Florida, Kentucky, and North Carolina. And plenty of northern states do poorly: Rhode Island, Connecticut, New Jersey, Minnesota, and Wisconsin."
It's probably worth taking into account that, in looking at the NAEP reading scores for the same grade (8th), the former group are all in the bottom half of the 50 states, and the latter group are all in the lower half.
IOW, the wider racial reading score gaps in (some) Northern states, compared to the same gaps in (some) Southern states, may have more to do with higher White achievement in the former group of states, than with higher Black achievement in the latter.
Generally speaking, greater equality can be achieved in two ways: by leveling up, or by leveling down. To me, the NAEP reading score data indicates that Southern states have "achieved" greater equality via the latter path.
How far ahead are the white kids from average?
Like, if all DC white kids are rich and have tutors they might be 50% above average. Vermont white kids might be behind.
This compares two numbers which can move and aren't being controlled for wealth disparity..
Vermont actually has some of the best educational outcomes overall. All three of my kids went to Vermont schools. They're well educated but depressed. Maybe they're depressed because they're well educated.
This raw data doesn't surprise me.
Vermont: not a lot of black kids, and those that are there are almost all in the Burlington area (~5K out of ~8K total), which is the richest area of Vermont. Not surprising at all.
DC: tons of black kids and de facto segregation, and lots of expensive private schools for the white kids of DC-living professionals to attend. DC has more black folks than white folks (~40% to ~38%) and is heavily segregated along racial/economic lines.
West Virginia is near the top?! You really should have "normalized" against the performance of the white students. If Oklahoma and West Virginia just has sucky schools then all the students will be clustered at the bottom and the differences by race would be compressed
"It's not clear what the differences are, but it's not much related to ancient patterns of slavery and racism."
Maybe segregation. The Black population in Wisconsin is about 6-7%, i.e., below the US average of about 14%. Wisconsin continues to be one of the most segregated states in the country. About 90% of the WI black population is concentrated in low income neighborhoids in a cluster of 5-6 counties in SE WI, and about 60% of Black people in Wisconsin live in Mikwaukee's lower income neighborhoods. White flight, historical redlining, plain old bigotry.
Compare the reading scores of white children in Alabama and Wisconsin and you'll see why there's a gap.