NAEP Test Scores: Mississippi Miracle vs. Oregon Outage

Which states do the best job of educating their public school students? Which do the worst?

Which states do the best job of educating their public school students? Which do the worst?

It’s easy to rank states by their raw scores on the federal government’s National Assessment of Educational Progress test, but the learning potential of public school students differs greatly between states. For example, averaging four data points — four and eighth grade 2024 reading and math scores — nationally, Asians average 269, whites 256, Hispanics 233, and blacks 229. Hence, as predicted by Moynihan’s Law of the Canadian Border, more northern states tend to score higher.

So, it’s easier for a state with a lot of Asians and whites to score higher than a state with a lot of blacks and Hispanics. Hence, it’s tricky to figure out which states are doing best at schooling their students, such as they are.

In case you are wondering about NAEP scores, just as the SAT is scored on a 200 to 800 scale with an intended mean/median of 500 and a standard deviation of 100, the NAEP is supposed to be scored on a 0 to 500 scale with a mean of 250 and a standard deviation of 35. But the NAEP is given to 4th graders, 8th graders, and occasionally 12th graders, and the older students score higher.

Practically nobody besides me averages the four 4th and 8th grade reading and math scores to come up with a single number to use in ranking states, but it certainly simplifies things.

Then, I’m going to use the Urban Institute’s methodology for adjusting NAEP scores for demographics (e.g., liberal Vermont has basically no blacks or Hispanics, so it ought to score considerably higher than nearly half-black Mississippi).

First, I’ve taken the raw NAEP scores released in January 2025 (4th grade reading and math and 8th grade reading and math) and calculatd a simple average to rank each state.

Here are the familiar raw score rankings unadjusted for demographics:

Average of 4th and 8th grades reading and math scores, unadjusted:

Not many surprises here: Massachusetts is the brainiest and/or best educated state, a title it has probably held on and off since the 1630s. On the other hand, Massachusetts’ NAEP scores are down to 256 in 2024 from 266 in 2013.

New Jersey and New Hampshire are in second, followed by some northern mountain states: Utah, Idaho, and Colorado.

New Mexico does quite badly in last place. The 25 point difference between Massachusetts and New Mexico is 5/7th of a (stylized) standard deviation, or 0.71 z.

One surprise is that Mississippi, so long the subject of “Thank God for Mississippi” exclamations from rival states struggling to stay out of last place on various measures, has lately reached the middle of the pack with a 245, for 29th place, ahead of much richer New York, Maryland, Texas, and California. Mississippi’s students are 43% white, 47% black, and only 1% Asian.

But these demographically unadjusted scores raises some old nature vs. nurture questions. Does Massachusetts do a good job of educating its kids, so other states should study how Massachusetts does it, or does it just have smart kids?

Not surprisingly, the public school district with the highest test scores in the country, according to Sean Reardon’s Stanford database, is Lexington, MA, where many Harvard and MIT professors choose to raise their families. But Lexington and Concord have also been the intellectual center of America since the Revolution started in Ralph Waldo Emerson’s grandfather’s backyard:

By the rude bridge that arched the flood,
Their flag to April’s breeze unfurled,
Here once the embattled farmers stood
And fired the shot heard round the world.

So, what’s the chicken and what’s the egg in Massachusetts?

And is Mississippi working wonders with what it has got?

And what state has the most self-destructive education policies?

From the Urban Institute:

States’ Demographically Adjusted Performance on the 2024 National Assessment of Educational Progress

Matthew Chingos, Kristin Blagg

January 29, 2025

Earlier today, the federal government released the 2024 scores from the National Assessment of Educational Progress (NAEP). NAEP is the only nationally comparable measure of student achievement that is reported for every state on a regular basis, but comparing states’ NAEP scores is misleading for many purposes because states serve very different student populations. For example, more than 20 percent of children live in poverty in Alabama and Mississippi, compared with less than 10 percent in New Hampshire and Vermont.

For nearly 10 years, the Urban Institute has published adjusted scores that capture how well students in each state score on the NAEP compared with demographically similar students around the country. We determine these adjustments by calculating how each individual student who takes the NAEP scores relative to students nationwide who are the same gender, age, and race or ethnicity and have the same free and reduced-price lunch receipt status, special education status, and English language learner status.

So, they largely adjust for race/ethnicity, but also throw in adjustments for some other demographic characteristics that the NAEP collects: income (as measured by free lunch eligibility), special ed, and immigrant kids who are learning English at school.

Seems reasonable.

We calculate adjusted scores for 2024 using our analysis of student-level data from 2022 and applying it to the unadjusted 2024 scores released earlier today.

In other words, to get their 2024 adjusted numbers out fast on the first day the unadjusted numbers were available, they just re-used the demographic adjustments they had previously laboriously calculated in 2022.

Sounds okay.

Our prior research has shown that this is a reliable way to adjust the state-level scores before student-level data are available.

These adjustments have important limitations, including challenges with accurately measuring income across different states and differences across states in the implementation of programs like special education.

For example, it could be that in some states, there’s a cultural tendency for, say, rich, smart parents to get their kids some kind of special education status so they get more time on tests or whatever.

But the adjusted scores come closer than the unadjusted scores to capturing the relative effectiveness of state policies.

That they do.

So, here’s my graph of their adjusted scores, averaging the four 2024 data points (4th and 8th grades, reading and math):

Average of 4th and 8th grades reading and math scores, unadjusted in blue, adjusted for demographics in red:

The Mississippi Miracle is number 1 after adjusting for its unpromising demographics, with comparable Louisiana in second.

The Mississippi Department of Education’s press release boasts of Mississippi’s striking improvement since the legislature passed a number of laws in 2013 (modeled on Florida’s 2002 reforms) to get serious about teaching reading and math.

Louisiana followed Mississippi’s lead, with comparable payoffs. Alabama has more recently followed its neighbors, which seems to have lifted Alabama’s test scores from awful to not-so-hot. Time will tell if Alabama follows its neighbors Mississippi, Louisiana, and Florida into the adjusted top ten.

I’ll discuss some of the other states first, then return to an in-depth exploration of the Mississippi Miracle and the Oregon Outage after the paywall.

Massachusetts comes in third in unadjusted scores, so it appears to be doing a good job on nurture as well as exploiting its nature.

I vaguely recall that around 1990 there was a broad movement in Massachusetts to upgrade the quality of government from the traditionally corrupt and lackadaisical kind seen in movies like The Departed to something more suitable for the state’s impressive level of talent. For example, Bill Bratton became top cop in Boston in 1993.

Texas comes in fourth when its mediocre demographics are accounted for. Texas used to put a thumb on the scale of its NAEP scores by encouraging more of its dimmest students to stay home on NAEP test days than most other states, but recent reforms have led to most states having participation rates within a narrow band. For example, on the 2022 NAEP, Texas was a little above average in getting both 4th and 8th graders to show up and take the test.

I’m still not sure I believe Texas’s NAEP scores, but Texans are not dumb. The state has mostly poor soil, so it never had a lot of farmers. And it had oil and gas rather than coal, which seemed to attract a better sort of worker than, say, West Virginia. (Old time pick and shovel underground coal mining was a hard, terrifying job, so people with other opportunities avoided it.) Ergo, Texas has, over the last 160 years, tended to attract people looking to get ahead in the modern American economy rather than in farming or underground mining.

Worst on the demographically adjusted list is Oregon, which has perhaps the highest percentage of truly Woke crazies in the country. Unadjusted, Oregon outscored the country on 8th grade math by 8 points in 2000, but trailed by 4 points in 2024.

Oregon 8th grade math, unadjusted:

How is Mississippi doing it? How is demographically blessed Oregon botching up so badly? Is there a correlation between how states vote and how well their schools work?

These questions are answered, to the best of my ability, in the 1,600+ words after the paywall.