Sunday, July 30, 2006

I'm leaving today for another summer vacation, but it will only be a week this time. (Only a professor would say my 2nd vacation in 6 weeks is only one week). But I will be back Sunday with all the usual GSS obsession and occasional smartassiness.

Saturday, July 29, 2006

Liberals say tax evasion is no biggie: Conservatives are the government haters just looking for any chance to get around paying taxes, while liberals seek the collective good by looking for ways to pay extra, right? According to the GSS, 43.7% of extreme conservatives say it's seriously wrong to cheat on your taxes. For extreme liberals, it's a whopping 10.9%. But let me anticipate some of you: it's always conservatives who lie on surveys.

Friday, July 28, 2006

Evidently, education still stinks for many blacks in the South: Steve Sailer pointed to improvements in education in the South as a possible reason for vocab gains among blacks in last few decades. So I looked at trends for 9 U.S. regions for both blacks and whites, and this is what I found:

Blacks--mean vocab score (out of 10) and % change since 1970s

1. West North Central 5.55 (+11.0)
2. Middle Atlantic 5.38 (+10.2)
3. West South Central 5.32 (+14.4)
4. East North Central 5.28 (+9.8)
5. South Atlantic 5.00 (-11.8)
6. Pacific 4.87 (-10.8)
7. East South Central 4.68 (-16.4)

(I excluded the New England and Mountain regions because of small sample sizes.) While the WSC region has seen big improvements (TX, OK, LA, AR) the rest of the South has actually gotten worse. The Pacific region also sunk 10.8%.

Whites--mean vocab score (out of 10) and % change since 1970s

1. New England 6.86 (+2.7)
2. Pacific 6.79 (+6.0)
3. Mountain 6.71 (+7.0)
4. West North Central 6.44 (+3.5)
5. Middle Atlantic 6.39 (-3.9)
6. South Atlantic 6.38 (+12.5)
7. East North Central 6.16 (+1.0)
8. West South Central 6.15 (+2.7)
9. East South Central 5.86 (+4.6)

For whites, New England is about to lose it's first place position to either the Pacific (or even Mountain) region. California has been losing a lot of working- class whites with modest vocabs. What is intersting is that the mountain states have been absorbing many of those migrants, and yet have improved slightly more than California has. The other big gainer has been the South Atlantic which includes DE, MD, and DC, along with all the other states you would think would go in this category.

One pattern I notice is that improvement has been more consistently positive for whites across the country (albeit modest in some places) while regions have varied more for blacks. Is this due to a greater standard deviation of environments for blacks?

Another thing I see (and the data from this decade are no different from what I have seen before) is that the dumbest whites and the dumbest blacks like to live in Kentucky, Tennessee, Alabama, and Mississippi. Is it somethin' in the cookin' down there? The problem is that, according to these data, the white folks there are moving up while the black folks are slipping farther behind. But whites there better not get cocky because if current trends continue, they could get passed up by blacks from places like Nebraska and Minnesota.

Wednesday, July 26, 2006

Perhaps the rising vocabs are due to more education, or both are due to increasing...something: A reader suggested as a straightforward explanation that whites and especially blacks are doing better on the GSS vocabulary test because they are staying in school longer. I calculated the mean years of schooling for both races for each of the past four decades to see if the pattern matches the vocab trends. They look to me like they do. Take a look:


Blacks--mean years of schooling
70s 10.2
80s 11.4
90s 12.3
00s 12.6

Whites--mean years of schooling
70s 11.9
80s 12.5
90s 13.3
00s 13.6

With vocab, we saw a steady improvement among blacks. We see that here too, although the rate of increase in education shrinks over time. With whites, they only saw improvements in knowledge of words between the 80s and 90s. Like blacks, their educational levels have improved every decade since the early 70s, but the biggest increase was between the 80s and 90s, the same span of time that they developed better vocabs.

In the past 3 or 4 decades, many more Americans are getting at least some college experience, and college is a place where many people hear words they rarely hear any place else--the very words that are likely to show up on a vocab test. This is especially true of blacks.

Or, something deeper is going on which has raised educational and word knowledge levels together? Have people gotten smarter, hungrier for learning because of better nutrition perhaps? This sounds more plausible if we're talking about the beginning of the century, not the end of it. Perhaps the growth of the "information society", which pushes people to stay in school and exposes them to more novel words? What do you think?

Tuesday, July 25, 2006

The "Flynn effect" observed in GSS data: Now we can't get too excited about measuring IQ with 10 vocabulary questions, but what the heck--this is a blog, not Intelligence. The GSS has been asking vocab questions since the early 70s, so I averaged the number correct out of 10 for blacks and whites. Here they are:

Blacks
70s 4.59
80s 4.69
90s 4.97
00s 5.15

Whites
70s 6.17
80s 6.14
90s 6.33
00s 6.39

So we see increases for both groups, but it's larger for blacks. While whites increased 4% over the period, black averages rose 13%. One bit of evidence that this measure is not just nonsense is that we see the consistent, large gap between blacks and whites, like a guzillion IQ studies have shown. And there has been some narrowing of the gap over the period.

Now it's possible that the words got easier--I don't know if they used the same words every year, and it's too late for me to look the damn thing up. I will say that if the test got easier, then why did whites only improve between the 80s and 90s while blacks improved every decade? And why did blacks improve more? I thought as people age, they might gradually pick up more words, and while it's true that mean white age rose 2.6 years over the period, blacks dropped .4 of a year. And whites aged gradually from the early 70s until now, while their vocabs only improved between the 80s and 90s. Comments are appreciated, and maybe I'll try to see what explains the increase later.

Sunday, July 23, 2006

According to the "foot vote," European immigrants say that New England is the best place to live in America: A reader brought up the question of desirable living places, as indicated by migration. While I cannot answer his specific questions (at least for now) I can look at the GSS to see where immigrants to the U.S. from wealthy European countries choose to live. The GSS only allows us to look at 9 regional divisions of the country, so I identified regions which had more European immigrants than would be predicted by total immigration patterns. Below I've listed the 9 divisions and the groups overrepresented there (and those in parentheses are underrepresented):

New England
French Canadians
Irish
Italians


Middle Atlantic
Italians
(English)
(French)
(Germans)

East North Central
Austrians
Dutch

West North Central
Germans

South Atlantic
English

East South Central
none

West South Central
Germans

Mountain
English
Scots

Pacific
(Italians)
(Austrians)
(French Canadians)

A couple of points: First, it looks like there is a tendency for immigrants from wealthy European countries to go regions where there are concentrations of their co-ethnics. We know that much immigration is family-based. Second, these immigrants are gravitating to whiter regions. New England is the whitest division, and it has 3 overrepresented immigrant groups and no underrepresented ones. East North Central and Mountain regions have 2 pluses and no minuses. The increasingly non-white Pacific region has 3 underrepresented groups. So judging by European immigration, New England is the most desirable place to live in America.

Friday, July 21, 2006

Mexican-Americans rank dead last on vocabulary, while Jews rank first--big surprise: The GSS asks respondents 10 vocabulary words. Below are the average number of correct answers for various American ethnic groups. I included only those born here since immigrants are at a disadvantage:

1. Jews 7.65
2. Chinese 7.17
3. Lithuanians (non-Jewish) 7.12
4. Austrians (non-Jewish) 7.00
5. English/Welsh 6.85
6. Danes 6.84
7. Scots 6.72
7. Yugoslavs 6.72
9. Russians (non-Jewish) 6.71
10. Norwegians 6.65
11. Swiss 6.64
12. Swedes 6.60
13. Hungarians (non-Jewish) 6.58
13. Finns 6.58
15. Japanese 6.54
16. Czechs 6.51
17. Irish 6.42
18. Greeks 6.41
19. Italians 6.40
20. Poles (non-Jewish) 6.36
21. French 6.30
22. Germans 6.28
23. French Canadians 6.19

U.S. average 6.17

24. Portuguese 6.13
25. West Indians 6.10
26. Rumanians (non-Jewish) 6.00
26. Arabs 6.00
26. East Indians 6.00
26. Belgians 6.00
30. Dutch 5.89
31. Spain 5.73
32. Filipinos 5.27
33. American Indians 5.20
34. Blacks 5.04
35. Puerto Ricans 4.99
36. Mexicans 4.83

Jews are in a class all their own here. Compared to Mexican-Americans, they answered 60% more of the questions correctly. They inflated substantially the means of several countries, so I excluded them where they did. Whites are generally in the top half, while poor minorities, especially those who use English less at home, drop to the bottom. Mexicans might work hard at some things, but not at building up a decent vocabulary. The exceptions to this pattern are Chinese Americans who are ranked second, and the Japanese who are above average (but significantly below the Chinese). The Dutch and Belgians are white groups who rank toward the bottom. Those from Spain are even lower--there is a tendency for southern Europeans to have poor rankings.

Comparing these numbers to national IQ estimates (Lynn's work) most of the rankings are similar, but a few are strange. Japan has the highest average IQ of any nation (except for South Korea, perhaps), but does not rank so high in this list. (Of course, NE Asians are strongest on visio-spatial skills, but why then are the Chinese at the top of my list?) Studies of Belgians and the Dutch estimate IQs around 100, but the numbers here are much lower. Spain is in the high 90s, but it ranks 31st in this list. (Some Hispanics might say they are from Spain). In my list, Greek Americans are above average, but according to Lynn, Greece is in the low 90s. Of course, some differences could be explained by immigrants and their descendants being different from the average back home. This might explain how West Indians score almost the U.S. average in terms of vocabulary. My numbers seem right for Filipinos since they have low average IQs, but that fact is at odds with all my other posts which show that Filipinos are high quality immigrants.

Thursday, July 20, 2006

Don't trust social science researchers: A colleague told me once that research shows that one's college experience does not affect political orientation. I told him I was skeptical of this idea since colleges live to crank out fresh liberals. I know most professors put people to sleep, but you mean to tell me four years of indoctrination does not work? Well, let's see what the all-wise, all-knowing GSS says. I looked at the question about your politics ranging from extremely liberal (1) through extremely conservative (7).
Here are the means for freshman through seniors:

Mean "conservatism" score:
Freshman 4.47
Sophomores 3.53
Juniors 3.46
Seniors 3.40

We see here a very clear move left as one accumulates years of college. The freshman average is between moderate (4) and slightly conservative (5), while seniors are between moderate and slightly liberal (3).

I recommend that you folks do as much data analysis of your own as you can because, frankly, I don't trust social science researchers.

Wednesday, July 19, 2006

There are more American Indian than Mexican voters: I looked at the 2000 presidential election (GSS data, like usual) to see the relative size of ethnic groups that voted. It looked like this:

Share of voters
1. Germans 16.9%
2. English/Welsh 14.7
3. Irish 14.3
4. Blacks 9.4
5. Italians 5.5
6. Scots 4.2
7. American Indians 3.9
8. Mexicans 2.8
9. Poles 2.6
10. French 2.0
11. Swedes 1.8
12. Norwegians 1.9
13. Russians 1.8
14. Dutch 1.3
15. Puerto Ricans 1.1
16. Filipinos 0.6
17. Chinese 0.6

These data support Steve Sailer's argument that Mexicans nationally are still a tiny group of voters. I can't remember the last time I laid eyes on an American Indian (AI), but even their numbers are substantially larger. If Republicans want to feel multicultural, why don't they go after a larger AI vote? After all, Republicans are like Indians--they are manly and aren't whiny, they love freedom, and tobacco. Ten percent of New Mexico is AI, and that is a swing state. In 2000, 56% of AIs voted for Bush, but only 46% voted at all, so why don't they have a "get out the AI vote?"

More later....

Monday, July 17, 2006

Smart people are politically more extreme: A reader hypothesized that smart people tend to have more extreme political views, and suggested a post on the educational levels of third party voters as a test of this. According to the GSS, Americans with graduate degrees were 12.1% of voters, but were 20.0% of Nader voters. While 2.5% of voters opted for Nader, 4.1% of those with advanced degrees voted for him.



Bush should thank lazy, illiterate liberals! I was interested in how true is the idea that Democrats lose votes because people sympathetic to their views are less likely to vote, and due to high school dropouts not voting as much. Here's the breakdown of the percent who voted in the presidential election in 2000:

Percent voting in 2000
Extremely liberal 60.2%
Liberal 61.8
Slightly liberal 67.2
Moderates 59.9
Slightly conservative 70.2
Conservative 74.2
Extremely conservative 70.5

Look at the lazy liberals! God bless 'em! (Well, maybe not in the case of Bush). They are no more likely to vote than moderates. I was told that liberals are the ones who truly care about democracy, but it's conservatives who actually behave like it's important.

Percent voting in 2000
less than high school 38.3%
high school 60.8
junior college 72.4
bachelor's 79.0
graduate degree 84.3

Wow, people with bachelor's degree (or more) are twice as likely to vote as high school dropouts. It does look like Democrats suffer a lot from liberals and illiterates not voting, but I'm not going to lose any sleep over people not casting ballots who can't spell "casting" or "ballot".

Tuesday, July 11, 2006

All you men out there, I've got news that will restore your self-esteem! Bonehead researchers studying penis length traditionally allowed participants to measure themselves. That's like allowing a student to grade himself. So the average finding was generally over 6 inches, and millions of men who feared they were inferior from watching porn felt it must be true if research said so. Well, one group of researchers had some sense and did the measuring themselves. The average from groin to tip was 5.1 inches. So gentlemen, cheer up! You're not as bad as you thought! Now, I know of you HBD guys have race on your minds, but I'll have to get back to you on the composition of the sample. (Wessells, H. et al. 1996. Journal of Urology 156:995)

Monday, July 10, 2006

A meta-analysis showing that blacks have higher rates of psychopathy: As Michael Levin has written, sometimes it sounds like the only difference between whites and blacks is IQ. I suspect it is more accurate to ask where aren't we different:


Putative ethnic group differences in various forms of psychopathology may have important theoretical, clinical, and policy implications. Recently, it has been argued that individuals of African descent are more likely to be psychopathic than those of European descent (R. Lynn, 2002). Preliminary evidence from the Psychopathy Checklist: Youth Version (A. Forth, D. Kosson, & R. Hare, 2003) offers some support for this contention, with Black youths rated as moderately higher (Cohen's d = .61) than White youths in a large institutional sample (n = 945). To examine this issue more exhaustively, the authors meta-analyzed adolescent psychopathy data from several studies (combined N = 2,199) and obtained a much smaller mean difference (dw = .20, p = .03), although considerable heterogeneity was evident among the effect sizes. (McKoy WK and JF Edens. 2006. J Clin Consult Psych.)

Sunday, July 09, 2006

More on dislike for blacks: As a follow-up to the previous post, a reader suggested that, on average, Jews might have less warm feelings toward blacks than Christians. I looked at the GSS and found that Christians, Jews, and those with no religion all had around 88% feeling warm or neutral.

As an aside, only 2.3% of white respondents said "very cool." Even if it is an underestimate, it is hardly supportive of the view that somewhere hidden under every white-bread suburbanite lurks a Klansman waiting to get out.

Saturday, July 08, 2006

Do conservatives dislike blacks? One vicious claim by liberals is that conservatives are racists. Is it true? I analyzed data from the General Social Survey and found that 81.25% of libs say they have warm or neutral feelings toward blacks (and, of course, pledging undying love for our darker brothers is mandatory, even if you are lying). So maybe half as many cons? Wrong--the actual number is 87.7%. (In the interest of full disclosure, cons had more neutrals, which is, incidentally, consistent with their belief in color-blindness).

Thursday, July 06, 2006

Diversity breeds violence

A quick glance at one of my university's periodical databases produced three studies showing that communities and nations with more ethnic diversity have more violence. This isn't the least bit surprising, but you'll never hear it talked about by the MSM as an important reason why the U.S. has a high homicide rate, or how perhaps diversity is a burden, not a godsend.


1. Osgood, Wayne et al. 2000. Criminology.

2. Roundtree, Pamela et al. 2000. Criminology.

3. Hoskin, Anthony. 2001. Justice Quarterly.

Consistent with the male predisposition for violence, they are more supportive of abortion on demand!



The National Organization for Women sells buttons that say, "If men got pregnant, abortion would be a sacrament." Yes, reproductive rights would be much more secure if it wasn't for misogynist men. The GSS has asked respondents since 1972 if abortion should be allowed for any reason if a women wants it. It may be difficult to see it in these graphs (men above, women below), but of the 19 years that the question was asked, women had a higher percent in favor compared to men exactly 3 times. Overall all the years, 41.2% of men support abortion on demand, versus 39.2% of women. And why wouldn't men support it more: they like sex without responsibility even more than today's women.

Wednesday, July 05, 2006

Political moderation shrinks as years of schooling grow



This graph shows the political orientation of people who completed various years of education. Starting from the bottom color and moving up, extreme liberals are disproportionately high school dropouts and overeducated eggheads. In addition, more years of education is accompanied by a greater share of liberals and slight liberals. Moderates are concentrated among the least educated, and are just barely the modal category among those with the most years of school. Conservatives and slight conservatives are found most often among the sensible, college-educated, but extreme conservatives are most often found among high school dropouts.

Tuesday, July 04, 2006

Blacks are most likely to say they have no ethnicity except "American": The General Social Survey indicates that almost all Americans say they belong to some ethnic group or groups. Only 2.5% say they are American and nothing more. This shows that ethnicity has not completely vanished from the American scene, but the interesting part is that of all the various factors I looked at, being black was by far the best predictor of saying that you do not identify with some group outside of the United States. Looking at the numbers below, 12.8% of them say this. Sixty-one percent of the people who say their identity is "American only" are black, even though blacks are only 12% of the U.S. population. Thirty-seven percent are white and only 2 percent of them are from the "other" racial category. So it seems that there is a small but significant share of blacks who feel no connection to Africa whatsoever.


Race
Blacks 12.8%
Whites 1.2
Others 1.6

Sex
Female 3.3%
Male 2.0

Degree
Less than high school 5.0%
Graduate 0.8

Occupational prestige
Low 4.2%
High 1.4

Income
Low 3.2%
High 1.3

Religion
Protestant 3.7%
Catholic 1.2
Jewish 0.7
None 1.9

Native-born
Yes 2.7%
No 0.5

Monday, July 03, 2006

Which group makes the best fathers? The growth in father absence is one of more alarming trends in recent American history. It has become the norm in black neighborhoods. So in this age of irresponsibility, who is doing the best job of getting and staying married when children are born? Here is the percent of fathers who are married for various American ethnic groups (from the GSS):

1. East Indians 95.5
2. Filipinos 94.4
3. French Canadians 90.0
4. Poles 82.1
5. English/Welsh 80.0
6. Dutch 74.5
7. Mexicans 74.4
8. Chinese 72.2
8. Scots 72.2
10. Irish 71.9
11. Italians 70.9
12. Germans 70.8
13. Norwegians 68.0
13. American Indians 68.0
15. French 65.4
16. Puerto Ricans 59.5
17. Blacks 48.7

Brown people are at the top and bottom of this list. Very few East Indian or Filipino fathers are not married, compared to half of blacks. Clearly, paternal commitment in the black community has collapsed. It survived Jim Crow just fine, but fell like a house of cards under liberalism. Puerto Rican men also have high rates of father absence. Chinese fathers are known for devotion to their children, but perhaps life in tradition-destroying cities like San Francisco have eroded this.