Bruno: American 17-Year-Olds Are Doing Better Than Ever

One of the reasons I like reading Kevin Drum is that he's one of the minority of pundits who, when talking about education, usually remembers that a lot of the news about American K-12 education is good.

This also means that when he tempers his edu-optimism, I stop and think.

So when Drum observes that while math and reading test scores over the years have improved "[q]uite a bit", this is true only "through 8th grade", I became curious about how high school outcomes have changed since the 1970s.

 If you look at the National Assessment of Educational Progress' "long-term trends" in math and reading, it's true that overall scores have been basically stagnant for 17-year-olds since the NAEP was first administered in the early 70's.

Below the fold, I'll explain why overall averages can be misleading and a deeper look into the data should brighten our outlook.

First, consider test scores by student race. As you can see in the graph above, since the first NAEP administration average reading scores for white 17-year-olds have risen barely at all: 4 points from 1971-2008. (The next set of results - from the 2012 administration - will be released this year.)

At the same time, reading scores have grown considerably faster for black and Hispanic students: 27 and 17 points, respectively.

We see a similar pattern in math. Between 1973 and 2008, scores for white students ticked up by a meager 4 points. Meanwhile, black students saw a gain of 17 points and Hispanic students' scores went up by 16 points.

Obviously, I'd love to have seen bigger achievement gains for white students and even bigger gains for students of color. And it's worth noting that many of the gains appear to have happened in the 70's and 80's, with gains slowing or stalling since then.

Still, it's important to note that overall scores look flatter than they are in part because of demographic changes, with lower-scoring groups of students becoming a larger share of the population. Most notably: from 1980 to 2008, Hispanics grew from 6% to 15% of the population, while white students fell from 80% to 66%.

Additionally, it's probably a mistake to look at high school achievement without also looking at dropout rates. Educational attainment is important in its own right and, while not as big of an issue in the lower grades, selective attrition has the potential to affect average test scores at the high school level.  

For example, in 1972 6.1% of students enrolled in high school dropped out. By 2008 that number had fallen to 3.5%. Improvements for striking for white students (5.3% to 2.3%), black students, (9.5% to 6.4%), and Hispanic students (11.2% to 5.3%).

Evidently, more students are persisting through high school. The presence of these additional (presumably) lower-scoring students should, if anything, be putting downward pressure on NAEP scores. Instead, scores have - at a minimum - held steady.

In other words, educational outcomes for students in the entire K-12 system are probably better than they have ever been. 

Describing educational achievement as "fairly flat" doesn't quite capture that. - PB (@MrPABruno)

Comments

[Bruce]

I think it's a mistake to confuse educational outcomes with achievement scores, or even attainment. Students' lives don't end at 17. My own evaluation is that, overall, we're doing better than we were 40 years ago, but the rate of improvement has been slower than the rate of increase in educational expenses, and our improvement has been inadequate to keep us ahead of certain faster-improving countries and regions of the world.

It's also always important to look for what is not immediately captured by easily available data. In your own subject field, science, America leads the world, in terms of prize-winning discoveries and patented inventions, but increasingly it is foreign-born talent that has been propelling us into these leads; while for Americans who have been here for many generations and who may wish to continue to lead only slightly modified 20th-century lifestyles like those they grew up with, they can feel the world around them changing, and they are worried about the future.

[Paul Bruno]

There are certainly other outcomes worth worrying about, but historically these outcomes - achievement and attainment - have been pretty well-correlated with many of those other outcomes. No doubt about it, though: I'd prefer faster improvements also.

I'm somewhat less worried about "competitiveness" internationally; improved educational outcomes in other countries are probably not a bad thing - particularly as other countries reap the low-hanging education fruit we reaped years ago - even for us (and certainly not for them).

Again, though, better is always better. I just think it's worth keeping in mind that improvements in American educational outcomes have been pretty substantial to date, given how large and complex a beast our system is.