Militarily fit-to-serve by state (round 2)

Last weekend, I created a "fit-to-serve" index by state. As I was constructing it, it felt like too much emphasis was being put on the percentage of each state's population on parole, on probation, in jail, or in prison. So I went back and reworked the numbers in to create a more straightforward, less arbitrary way of measuring eligibility by state--I simply added the totals of the three inhibiting factors together and then subtracted that sum from 100 to get the percentage of each state's young adult population that is deemed potentially fit-to-serve in the military*:

State	Eligible %
1. Vermont	59.8
2. Minnesota	59.2
3. Wisconsin	57.4
4. Iowa	57.1
5. North Dakota	55.4
6. Connecticut	53.0
7. Montana	52.7
8. Utah	52.4
9. New Hampshire	51.9
10. South Dakota	51.5
11. New Jersey	50.1
12. Pennsylvania	49.4
13. Maine	48.8
14. Missouri	48.2
15. Oregon	48.0
16. Wyoming	47.4
17. Maryland	47.3
18. Massachusetts	46.8
19. Nebraska	46.7
20. Colorado	46.6
21. Idaho	46.4
22. Kansas	46.1
23. Oklahoma	45.6
24. Rhode Island	44.2
25. Hawaii	42.9
26. Michigan	42.3
27. Ohio	42.0
28. Virginia	41.8
29. Washington	41.7
30. Illinois	41.4
31. West Virginia	40.5
32. Indiana	40.2
33. California	37.2
34. Kentucky	36.1
35. Arizona	36.0
36. Texas	35.5
37. Delaware	35.2
38. New York	34.1
39. Tennessee	33.5
40. Arkansas	32.6
41. North Carolina	32.4
42. Alaska	32.2
43. Florida	28.8
44. Alabama	27.9
45. New Mexico	23.1
46. South Carolina	22.4
47. Louisiana	21.2
48. Georgia	19.3
49. Mississippi	17.4
50. Nevada	15.9
51. District of Columbia	15.2

What immediately jumps out is how white (and geographically concentrated in the upper Midwest and Northeast) the states with high eligibility are compared to those with more modestly sized eligible populations. The correlation between the percentage of a state's population that is white and the percentage of the young adult population deemed eligible for military service is .67 (p = 0).

So much for the idea of granting citizenship to immigrants upon some set duration of military service--unless they are among the sliver of newcomers hailing from Europe, they won't be able to get in! Because far less than 1% of the adult population is actually serving in the military at any given time, to assert that the perpetual decrease in the proportion of the country's population that is white will make it difficult for the military to find potentially eligible recruits actually isn't justified (though being able to find willingly eligible recruits is a separate issue).

The report's public authors, a cadre of retired military officers, do not mention the demographic angle (nor should they, as it is of course irrelevant!). They do emphasize the putative benefit of early education in reducing criminality and increasing the likelihood of on-time graduation throughout, though--in fact, the report is subtitled "Early Education across America is Needed to Ensure National Security"!

Then, without any apparent sense of self-defeat, the report's appendix includes a table showing the percentages of each state's 4-year-old populations enrolled in pre-kindergarten schooling. This measure inversely correlates with latent graduation rates at a statistically insignificant .09 (p = .53) and probation or incarceration rates at a similarly statistically insignificant .06 (p = .69). That is, early education is not associated with desirable social outcomes like on-time high school graduation and steering clear of the law at the state level, despite the praise heaped upon early education and its supposed long-term benefits.

In the comment thread of the previous post, Silly Girl, perspicaciously detecting the lack of any relationship just by eye-balling the table, remarked:

Ugh, page after page of that report trumpeting the benefits of pre k education, then on page 7 the charts of pre k ed. and graduation rates showing no relationship to graduating based on going to free public pre k. Do they think people can't read and think? Okay, dumb question. Do they think no one, even the more educated folks who likely would read it, would question it?

Kurt9 didn't see anything intentionally furtive going on:

People in bureaucracies don't think, period. Its not that they thought they could slip this report past readers without them reading it critically. Its that they did not even think about this at all.

Whatever the explanation, it is, in the literal sense of the word, ridiculous.

Parenthetically, reassuring me that this method is superior to the index I previously created, the correlation between estimated average IQ and the percentage of the young adult population deemed eligible for military service is .78 (p=0) (IQ correlated with my "fit-to-serve" index at .54); as always, desirable social outcomes and intelligence go hand-in-hand.

* This method assumes no overlap among the three inhibiting factors, even though there likely is a significant amount of it--I suspect failing to graduate from high school on time, being overweight, and scuffling with the law all correlate positively with one another. But I see little reason to suspect the amount of overlap varies significantly by state--that they are likely correlated (the report from which the data come does not attempt to tease out what the true percentage of each state's population deemed eligible for service actually is) means the table above inflates the percentages of state populations deemed ineligible for military service, but in a systematic way that doesn't materially effect some states more than others.