Tuesday, May 11, 2010

State IQ estimates (2009)

With the 2009 NAEP scores out, it seems as good a time as any to update the IQ estimates by state that are based on the federally administered tests. Previously, I used math and science scores from 2005 to construct the estimates. However, science testing was not conducted in 2009, so I've substituted reading for science this time around.

The scores for both tests are on a 500 point scale, with a designed standard deviation of 50. In the proceeding table, these are are converted to IQ estimates with a mean of 97.5, corresponding to the national average NAEP scores of 281.67 for math and 262.29 for reading, and a standard deviation of 15. The math and reading scores are given equal weighting.

I'm operating under the presumption that the average IQ in the US is 98. The reason the IQ estimates are based on a national mean of 97.5 is to take into account private schools, whose students on average score 15 points higher on the math test and 19 points higher on the reading test, corresponding to an estimated 5.1 IQ advantage over their public school counterparts. Representing a little more than one-tenth of the school age population, this yields a national average of 98. The estimates assume that the private school advantage relative to public schools is equal across states. That is, the public school score, plus 5.1 multiplied by the percentage of a state's students enrolled in private schools, gives the state's overall estimated average IQ.

Again in the spirit of the 2004 IQ hoax, states are colored in accordance to how their populations voted in the 2008 Presidential election. Light red (blue) indicates the margin of victory for McCain (Obama) was less than 10%; dark red (blue) indicates it was 10% or more:

2.New Jersey101.4
5.New Hampshire100.9
8.North Dakota100.5
10.South Dakota100.3
27.New York98.7
33.North Carolina97.8
36.Rhode Island97.3
37.South Carolina97.0
46.West Virginia94.9
49.New Mexico94.8
51.District of Columbia91.2

* "Department of Defense Education Activity", comprised of the children of service members serving outside of the US. Despite a demographic pool (58.1% white, 22.0% black, 9.9% Hispanic, 9.0% Asian/Pacific Islander, 1.0% Native American) that suggests the DoDEA should be found near the bottom of the rankings, if measured as though an independent state, DoDEA schools place 11th, between Maryland and Wisconsin. Military men aren't morons.

An aunt of mine who lives in Alabama has joked that the state's motto is "thank God for Mississippi". The demographic transition the nation is undergoing, however, has already made her remark a bit antiquated. I suspect in a few years' time, both Alabama and Mississippi will be able to thank him for the Southwest.

Although I have substituted reading scores for science scores in converting NAEP scores into IQ estimates four years down the road, the 2005 and 2009 state level results still correlate with one another at a very vigorous .91 (p=0). There is a great deal of consistency in the average state scores among different cohorts of test takers, suggesting that deriving IQ estimates from NAEP testing is sound.

As has historically been the case, the further north one goes, the smarter the population becomes. During the 2008 election cycle, Half Sigma frequently decried the perceived abandonment of the GOP by the cognitive upper crust of the voting population. Indeed, the best-performing red states are only tenuously so. Middling Kansas tops the list of solidly Republican states. I suspect, however, given the staggering growth in government spending that is coming down the pike over the next four years, that 2010 will reveal a reversal in that trend to be occuring.


  1. Texas does almost as well as Utah. That's worth thinking about.

  2. A simple observation WRT the DoD:

    The US Military education and intelligence standards are very high and have been since the late 1970's when the US Army went all volunteer.

    Junior level officers are bachelor degree college graduates for the most part. Most upper level field grade military officers have master's degrees, senior grade officers have either multiple masters degrees or doctorates.

    Today's military Sgt Majors are better educated and have better miliary professional skills than the generals who ran WW2 Army Divisions and USAAF bomber Heavy Groups.

    The American Military can and does test for IQ in a way that is illegal under various civil rights laws for any other large private or American government entity.

    Since most American long service term professional officers and term non-commissioned officers come from multi-generation military families, there are a number of implications the HBD advocates to consider.

    Note as well that 80-90% of long term of service military officers are Republicans.

    Only university professors are as "partisan by profession" as military officers.

  3. "Only university professors are as "partisan by profession" as military officers."

    Hm, interesting idea.

    Where are all the PhD's?

    I wonder what percent of PhD's are in industry, education (k-12), academia (universities, foundations), military, clergy etc. and how they come out on the political spectrum/compass.

  4. Any idea of what the Canadian IQ is?

  5. I think an important factor in voting is that whites tend to be more conservative in areas with the most blacks and/or Hispanics and more liberals in areas with the fewest blacks and/or Hispanics. This may explain why Republican areas tend to be less intelligent than Democratic areas since blacks and Hispanics tend to be more common in the south than they are in the north.

  6. Mil Tech Bard,

    Right, the Department of Defense uses a categorization system, and the AQFT portion is essentially a set of IQ tests. In the last few years, the army has upped the percentage of Cat IV it allows in, I believe, but it's still a small number, and I don't think Cat V gets in at all. If memory serves from some of Steve Sailer's writing on it, the virtual minimum IQ for military personnel is 92.


    Not from NAEP, although Richard Lynn has an answer for you.

    Average Joe,

    Aargh, I can't find the post, but I know I've looked at it at the state level in the US, and you're correct, especially with regards to the size of the black population. That is, the more blacks a state has, the more Republican that state's white electorate is. I'll see if I can dig it up.

  7. >In the last few years, the army
    >has upped the percentage of Cat
    >IV it allows in, I believe, but
    >it's still a small number, and I
    >don't think Cat V gets in at all.
    >If memory serves from some of
    >Steve Sailer's writing on it, the
    >virtual minimum IQ for military
    >personnel is 92.

    That is a "kinda sorta Cat IV" and the AQFT is only a part of what is going on.

    The US Army has developed a number of HR recruiting and training tools to identify intelligent or highly motivated, but poorly educated by public schools, male recruits.

    James Dunnigan has spoken of this over on strategypage.com several times since 2005.

    The gist is that the PC anti-male bias of the largely women run large urban and older inner suburban public schooling has a lot of young males bailing on both high school and college.

    The result is the shrinking pool of male high school and college graduates has the Army going back to those men as recruits.



    In the last decade, the army has made a lot of progress in improving how accurately it screens risky candidates. It's not just the improved selection process, but improved training methods as well. A lot of high school drop outs were poorly served by bad urban schools. Similarly, many of those with criminal records had already put that sort of thing behind them, and were looking to the army for a new beginning.

    Until recently, less than ten percent of army recruits had been high school dropouts. But in the last decade, that has grown to 24 percent, with no noticeable decline in the quality of troops. Same thing with those receiving "moral waivers" (for having a police record). That has gone from 4.6 percent four years ago, to 6.2 percent. The more accurate recruitment and training methods are also widely used in the civilian sector, where employers don't want to miss out on any potentially good workers.

  8. A lot of those states at the top of the list actually benefit from a double whammy on the NAEP: (1) They have very few blacks bringing down the average. (2) Because there are so few blacks, a lot more white students go to public schools instead of private ones. For example, I grew up in VT and almost no one went to private schools since the public schools were mostly just as good.

  9. Stopped Clock5/15/10, 1:45 PM

    Did you want



    I remember it because I posted on it. If there's another post like that with other info, Im noDid you want



    I remember it because I posted on it. If there's another post like that with other info, Im not aware of it.t aware of it.

  10. Stopped Clock,

    It must've been before that post, because I can tell by my writing that I'd already looked at it empirically (that is, I basically alluded to it).

  11. The #1 state on the list, Massachusetts, elected Scott Brown for US Senate and #2, New Jersey elected Chris Christie for Governor.

    Positive trend indeed.

  12. Anon,

    It will be interesting to compare how far states shift rightward in the 2010 elections (relative to 2008) with their estimated average IQs.

  13. Attempting to correlate standardized test scores (IQ) to state voting patterns is a gross simplification. There was a study where respondents self identified as conservative-moderate-liberal and Democrat-independent-Republican.

    The two most similar groups demographically were conservative Republicans and liberal Democrats. Those two groups had the highest average level of education, and also the highest estimated IQ scores based on performance on a brief test included in the study. The conservative Republicans actually performed very slightly better on the exam.

    Essentially, if a white American exhibits educational attainment and income levels associated with liberal Democratic voting in Minnesota or Massachusetts, it is extremely likely that the same person would vote as a conservative Republican had they been born and raised in Utah or South Carolina by parents from that state.

  14. chg,

    Right, it needs to be made clear that between-state variation and within-state variation are not the same (I am aware of it). This theme runs throughout Andrew Gelman's Red State, Blue State book as well.

  15. made use of ur data here a while ago, btw. thnx!


  16. I see the D of C is the lowest. I assume this was computed while Congress was in session, the Supremes seated and the President in the White House?

  17. Isn't it embarrassing that only ten US states have higher average IQ than the globally accepted average of 100?

  18. Anon,

    The global average is closer to 90. 100 is the average for whites of northwestern European descent.

  19. Rankings on educational attainment should not be mislabeled as IQ.

  20. Teri,

    There is nothing misleading in the post, and the correlation between scholastic aptitude and IQ is both strong and well-established. If you care to dispute the methodology, do so in a manner that actually shows you have some idea what you're talking about. Anyone can make unsubstantiated blanket statements with an air of gravitas when there is in fact nothing there at all.

  21. What frightens me is that I had always heard that the average IQ in the US was 100 exactly. When you see the breakdown by state and see how many points off from that idealized average they are it makes me worry that the US is rapidly getting dumber.

    I chuckled when I saw DC's IQ.

  22. Any wonder that 15 of the top 20 states voted for Obama over McCain? None that I can see. And fits my sense of the parties overall.

  23. Lots of Knee Grows in the stupid states!