Saturday, December 04, 2010

Winter wights and helio humans

Using the Derb's Ice People/Sun People dichotomy, the following table ranks states by their net frigidity, determined by taking a state's rate of change in Ice People minus its rate change in Sun People, as reported by Pew over the period 2005-2007, due exclusively to internal migration. It is necessary to account for differences in national population size, so the Ice People figures are based largely on white movement (93% white, 7% Asian). For Sun People, it is 55% Hispanic, 45% black*:

Frost factor
1. District of Columbia
2. Alaska
3. Maine
4. Kentucky
5. Montana
6. South Dakota
7. Wyoming
8. Louisiana
9. Arizona
10. Rhode Island
11. Vermont
12. New York
13. New Mexico
14. Texas
15. Illinois
16. Alabama
17. Michigan
18. Colorado
19. Virginia
20. Florida
21. Wisconsin
22. Mississippi
23. Oregon
24. California
25. Idaho
26. New Hampshire
27. North Carolina
28. Ohio
29. Washington
30. West Virginia
31. New Jersey
32. South Carolina
33. Nebraska
34. Massachusetts
35. Georgia
36. Indiana
37. Oklahoma
38. Delaware
39. Maryland
40. Pennsylvania
41. Tennessee
42. Connecticut
43. Utah
44. Kansas
45. Nevada
46. Minnesota
47. Arkansas
48. Missouri
49. Iowa
50. Hawaii
51. North Dakota

The nation's capital is a glacial fortress, with more than 3,000 Ice People moving in each year while 14,000 Sun People leave in search of sunnier settings. Parenthetically, keep in mind that the index is measuring net migration of Ice People relative to that of Sun People. The iciest places like DC, Alaska, and Maine are all losing people through internal migration, but they're losing Sun People faster than they are losing Ice People (or in the case of DC--the starkest example of white people going in one direction and NAMs going in the other--losing more Sun People than they are gaining Ice People), while top destinations like Idaho show up in the middle of the rankings because people of all races are heading there.

Other states (Alaska, Maine, Montana, South Dakota, and Wyoming) known for being cold prior to the Derb's pessimistic prognostications also tend to be relatively more attractive to Ice People than they are to Sun People. Minnesota, already renowned for its sizable Somali population in Minneapolis, is an exception to this cool rule. Whites and Asians steadily moved out over the period while Hispanics and blacks moved in. North Dakota is another exception, and an apparently anomalous one at that. Pew reports that the state's Hispanic population grew by over 8% annually from internal migration alone, which stands out along with Alaska (where the Hispanic population is reported to have declined 11% per year) as being far higher than Hispanic movement elsewhere.

Steve Sailer has noted that Hispanics tend to vote like whites do, shifted to the left several points. They also tend to move to and from the same places (Hispanics already resident in the US, that is--even though there were hundreds of thousands of immigrants from Latin America pouring into California over the period measured, nearly 300,000 Hispanics already living in the US left California during that time). The correlation between the white and Hispanic rates of change through internal migration at the state level is .45. For whites and blacks, there is no relationship whatsoever, while for whites and Asians it is .21 but falls outside of statistical significance at 90% confidence.

Going through the state data, it becomes clear that blacks don't move across state lines much, while Hispanics are quite mobile. The average absolute rate of change for blacks is 1.1%. It is 1.7% for whites, 5.3% for Hispanics, and 5.8% for Asians (although the high Asian figure is probably an artifact of the way the estimates were made, in increments of 1,000--in states like Montana and Wyoming, a movement of 1,000 Asians nearly doubles or halves the population, depending on which way they're heading). In light of this, it is not surprising that black urban culture emphasizes city of origin as an important marker of identification, manifesting itself in clothing, tattoos, and in the world of hip hop.

* Using US Census data, non-Hispanic whites constitute 65.1% of the US population, Asians 4.6%. 65.1/(65.1+4.6) = .934003, hence the 93% figure (index values are rounded after all calculations have been made). Hispanics comprise 15.8%, blacks 12.9%. 15.8/(15.8+12.9) = .550523, hence the 55% Hispanic figure.


Saint Louis said...

"The nation's capital is a glacial fortress, with more than 3,000 Ice People moving in each year while 14,000 Sun People leave in search of sunnier settings."

As shown by looking at Maryland's ranking of 39. There are places just a few blocks east and south of Capitol Hill that white people wouldn't have wanted to drive through, let alone live in, just 10-15 years ago. Now many of these places have been recently gentrified. As gentrification has pushed east and south from the Capitol, blacks have moved farther east and south, many crossing into Maryland.

Many first-line suburbs around the country are dealing with new problems thanks to the gentrification of inner city neighborhoods pushing poor blacks outward.

Audacious Epigone said...

Saint Louis,

The data backs your assertions up. Maryland had the country's ninth highest rate of growth in black population, suggesting that a lot of DC's blacks are leaving the nation's capital and heading there.

ray said...

Blacks don't move much any more because their households are female lead. A woman will not tend to leave the support of mom and grandma in search of a better job. Most black moms are single moms who are dependent on other family for childcare. etc. Back in the forties and fifties when the black illegitimacy rate was much lower a great migration of blacks took place from the sharecropper south to the factory jobs of the north. Welfare is a blessing in the sense that it keeps black families from moving much.

The Undiscovered Jew said...

Steve Sailer has noted that Hispanics tend to vote like whites do, shifted to the left several points.

Have you noticed the Hispanics (such as Marco Rubio, Brian Sandaval, Quico Canseco, Raul Labrador, etc) who were elected in 2010 as Republicans are the most Iberian looking of Hispanic Americans?

I'd bet that, to the extent the GOP is getting any Hispanic votes, it is coming from the 30% of Hispanic Americans who are either European (7/8 white or higher) or mostly European (>=3/4).

If GOP wants to get Hispanics to vote Republican they should raise the educational requirements legal Hispanics need to immigrate to America in order to attract better educated (read: whiter) Latin Americans who are more likely to vote Republican.

The pool of white Hispanics to choose from is not neglible either. Depending on how many white Hispanics are living in Southern Brazil, there are between 130-150 million white Hispanics in Latin America for our immigration policy to select from.

Not that I expect my ideas to be implemented, mind you.

Audacious Epigone said...


Astute observation. Ideally the GSS would have some question for Amerindians similar to the one for blacks re: complexion, but I don't expect that one to be implemented, either (nor the black one ever again--the last time it was before I was even born, in 1982).

Audacious Epigone said...


Lousiana came in second only to DC in terms of the percentage of its black population that it lost (actually, Maine falls b/w the two by Pew's methods, but that's likely a statistical artifact because ME's black population was so small to begin with) over the three year period. If blacks don't move much, big events like that look especially disruptive. Parenthetically, Texas looks to be the lucky recipient--about half a million blacks left LA during the three year period and ~450,000 entered Texas (both numbers are net).

The Undiscovered Jew said...

Astute observation. Ideally the GSS would have some question for Amerindians similar to the one for blacks re: complexion, but I don't expect that one to be implemented, either

AE, according to this*, among Hispanics there is a correlation between self reported ancestry and actual admixture levels:

* Admixture, ethnicity, and pigmentation of Hispanics and Native Americans

It was found that Native Americans tend to overestimate their genetic Native American-ness, while Hispanics tend to underestimate it. I don't find this surprising since Hispanics are descended from Spanish-influenced groups with Spanish as their language, and Catholicism as their religion, who viewed themselves in counterdistinction to the non-Spanish-influenced populations on the one hand, and English-speaking Americans to their north on the other. Thus, they may emphasize their "Spanish-ness" since it distinguishes them from both groups. Native Americans, on their other hand, view themselves in counterdistinction to Anglo-Americans, and identify with their native traditions, and hence tend to emphasize their "nativenesss".

Another interesting finding is that ethnicity among Hispanics (i.e. labels such as "half-white, half-Hispanic", "Spanish", "Mexican American" or "Mexican") is correlated with Native American genetic admixture in the expected order.

Skin color was also associated with genetic admixture, although the correlation wasn't as strong, and skin color was a relatively weak predictor of ancestral proportions. That does not argue so much against physical anthropological determination of race, but rather against the use of a single trait. Mark Shriver's group had looked at the correlation between skin pigmentation and ancestry in a previous study.


Results reveal that NAs underestimate their degree of EU admixture, and that Hispanics underestimate their degree of NA admixture. Within Hispanics, genetic-marker estimated admixture is better predicted by forehead skin pigmentation than by self-estimated admixture. We also find that Hispanic individuals self-identified as half-White, half Hispanic and Spanish have lower levels of NA admixture than those self-identified as Mexican and Mexican American.