The Myth of American Meritocracy: How corrupt are Ivy League admissions?

Via Sailer. Note: the article is by Ron Unz, though in this case his numbers appear consistent with my own impressions and previous knowledge.

The Myth of American Meritocracy:

The evidence of the recent NMS semifinalist lists seems the most conclusive of all, given the huge statistical sample sizes involved. As discussed earlier, these students constitute roughly the highest 0.5 percent in academic ability, the top 16,000 high school seniors who should be enrolling at the Ivy League and America’s other most elite academic universities. In California, white Gentile names outnumber Jewish ones by over 8-to-1; in Texas, over 20-to-1; in Florida and Illinois, around 9-to-1. Even in New York, America’s most heavily Jewish state, there are more than two high-ability white Gentile students for every Jewish one. Based on the overall distribution of America’s population, it appears that approximately 65–70 percent of America’s highest ability students are non-Jewish whites, well over ten times the Jewish total of under 6 percent.

Needless to say, these proportions are considerably different from what we actually find among the admitted students at Harvard and its elite peers, which today serve as a direct funnel to the commanding heights of American academics, law, business, and finance. Based on reported statistics, Jews approximately match or even outnumber non-Jewish whites at Harvard and most of the other Ivy League schools, which seems wildly disproportionate. Indeed, the official statistics indicate that non-Jewish whites at Harvard are America’s most under-represented population group, enrolled at a much lower fraction of their national population than blacks or Hispanics, despite having far higher academic test scores. [. . .]

Just as striking as these wildly disproportionate current numbers have been the longer enrollment trends. In the three decades since I graduated Harvard, the presence of white Gentiles has dropped by as much as 70 percent, despite no remotely comparable decline in the relative size or academic performance of that population; meanwhile, the percentage of Jewish students has actually increased. This period certainly saw a very rapid rise in the number of Asian, Hispanic, and foreign students, as well as some increase in blacks. But it seems rather odd that all of these other gains would have come at the expense of whites of Christian background, and none at the expense of Jews.

Furthermore, the Harvard enrollment changes over the last decade have been even more unusual when we compare them to changes in the underlying demographics. Between 2000 and 2011, the relative percentage of college-age blacks enrolled at Harvard dropped by 18 percent, along with declines of 13 percent for Asians and 11 percent for Hispanics, while only whites increased, expanding their relative enrollment by 16 percent. However, this is merely an optical illusion: in fact, the figure for non-Jewish whites slightly declined, while the relative enrollment of Jews increased by over 35 percent, probably reaching the highest level in Harvard’s entire history. Thus, the relative presence of Jews rose sharply while that of all other groups declined, and this occurred during exactly the period when the once-remarkable academic performance of Jewish high school students seemed to suddenly collapse. [. . .]

Each year, the Ivy League colleges enroll almost 10,000 American whites and Asians, of whom over 3000 are Jewish. Meanwhile, each year the NMS Corporation selects and publicly names America’s highest-ability 16,000 graduating seniors; of these, fewer than 1000 are Jewish, while almost 15,000 are non-Jewish whites and Asians. Even if every single one of these high-ability Jewish students applied to and enrolled at the Ivy League—with none going to any of America’s other 3000 colleges—Ivy League admissions officers are obviously still dipping rather deep into the lower reaches of the Jewish ability-pool, instead of easily drawing from some 15,000 other publicly identified candidates of far greater ability but different ethnicity. [. . .]

The situation becomes even stranger when we focus on Harvard, which this year accepted fewer than 6 percent of over 34,000 applicants and whose offers of admission are seldom refused. Each Harvard class includes roughly 400 Jews and 800 Asians and non-Jewish whites; this total represents over 40 percent of America’s highest-ability Jewish students, but merely 5 percent of their equally high-ability non-Jewish peers. It is quite possible that a larger percentage of these top Jewish students apply and decide to attend than similar members from these other groups, but it seems wildly implausible that such causes could account for roughly an eight-fold difference in apparent admissions outcome. Harvard’s stated “holistic” admissions policy explicitly takes into account numerous personal characteristics other than straight academic ability, including sports and musical talent. But it seems very unlikely that any remotely neutral application of these principles could produce admissions results whose ethnic skew differs so widely from the underlying meritocratic ratios.

One datapoint strengthening this suspicion of admissions bias has been the plunge in the number of Harvard’s entering National Merit Scholars, a particularly select ability group, which dropped by almost 40 percent between 2002 and 2011, falling from 396 to 248. This exact period saw a collapse in Jewish academic achievement combined with a sharp rise in Jewish Harvard admissions, which together might easily help to explain Harvard’s strange decline in this important measure of highest student quality. [. . .]

It is important to note that these current rejection rates of top scoring applicants are vastly higher than during the 1950s or 1960s, when Harvard admitted six of every seven such students and Princeton adopted a 1959 policy in which no high scoring applicant could be refused admission without a detailed review by a faculty committee.78 An obvious indication of Karabel’s obtuseness is that he describes and condemns the anti-meritocratic policies of the past without apparently noticing that they have actually become far worse today. An admissions framework in which academic merit is not the prime consideration may be directly related to the mystery of why Harvard’s ethnic skew differs in such extreme fashion from that of America’s brightest graduating seniors. In fact, Harvard’s apparent preference for academically weak Jewish applicants seems to be reflected in their performance once they arrive on campus.79

Related: "Merit" in elite college admissions

Racial Intermarriage and Household Production

For the edification of Whiskey and SeanTodRoy (pdf):

ASHG 2012 twitter copy/paste

Charley Farley ?@charley_farley Very cool PoBI talk from Leslie. I'm one of the red squares - Anglo-Saxon/pre-Roman British admix. Incredible UK structure #ASHG2012

Nick Eriksson ?@nkeriks Lovely talk by Stephen Leslie about fineSTRUCTURE and the fine structure of UK populations. learned lots about UK history. #ASHG2012

Luke Ward ?@luke__ward M.Wilson Sayres: Ultralow Y diversity due not only to diff reproductive success of sexes but also selection, both coding+noncoding #ashg2012

Yaniv erlich ?@erlichya WS: selection on the Y chromosome does not only act on the coding regions. #ASHG2012

Yaniv erlich Yaniv erlich ?@erlichya WS: reproductive success difference between male&female is not enough to explain the genetic signal from the Y chromosome #ASHG2012

Yaniv erlich ?@erlichya I am exited to hear Wlson Sayres talk about the Y chromosome! One of my favorite chromosomes #ASHG2012

23andMe ASHG 2012 poster presentations

Here. From "Genome wide association study of sexual orientation in a large, web-based cohort" (pdf):
Our GWAS results did not identify any genetic loci reaching genomewide significance at p < 5 x 10-8 among men or women. Among men, the peak (non-significant) hit was in chromosome 8q12.3 (chr8:63532921 in NKAIN3, p= 7.1 x 10-8).
More interesting (if generally unsurprising) are the phenotypic associations:
We examined the correlation between sexual identity and ~1000 phenotypes already characterized in the 23andMe database through other surveys. These analyses are preliminary; we have not checked for outliers or confounders beyond what is listed in the methods. We replicated previous findings showing a positive association between lesbians and alcoholism, and between lesbians and gay men and several psychiatric conditions.
A commenter at the 23andMe blog:
The phenotypic information is interesting if I’m reading it correctly: Gay men are less likely to to have played common US sports, and are more likely to cry easily or to have had liposuction. Lesbians are less likely to shave their legs. Surprisingly, gay men are less likely to be atheist or agnostic.

"New" R1a1 SNPs

New Y-chromosome binary markers improve phylogenetic resolution within haplogroup R1a1:

Despite the limited data available for Z280 and Z93, some general inferences can be drawn from the geographic distributions of these two haplogroups. The R1a1- Z280 subclade is a strong candidate for covering the R1a1a* (xM458) in Eastern Europe, which was found in high frequency by Underhill et al. (2010).The tested set of 53 Malaysian Indian samples presented 100% frequency for the R1a1-Z93 subclade, without co-existence Z280 or M458 sub-haplogroups. Inner and Central Asia seem to be the overlap zones for the R1a1-Z280 and R1a1-Z93 chromosomes as both forms were observed at low frequencies. This is again consistent with the observations described for R1a1a* spread in Central Asia and in the Altai region by Underhill et al. (2010). This pattern suggests that the origin of R1a1-M198 arguably occurred somewhere between South Asia and Eastern Europe. Potential candidates could be the Eurasian Steppes (Ukraine – Southern Russia – Kazakhstan – Caucasus) or the Middle East. European populations showed higher M458 and Z280, whereas Asian populations presented higher Z93 frequencies, indicating that the new markers can be effectively used to distinguish between the European and Asian branches of the haplogroup R1a1-M198. [. . .]

The coalescent time calculated by us for R1a1-M458 carriers is consistent with the age calculated by Underhill et al. (2010) in Europe yielding 7.3 KYA versus 7.9 KYA (thousands of years ago). Underhill et al. (2010) also noted the potential association of R1a1-M458 with the Linear Pottery Neolithic culture in the territory of present-day Hungary—this observation is supported by our data. The TMRCA calculated for R1a1-Z280 diversification (10.3 KYA) is approximately in agreement with the estimation of Underhill et al. (2010) for R1a1a*(xM458) chromosomes in Eastern Europe ( 11 KYA). However, the coalescent age of 10.3 KYA for R1a1- Z93 chromosomes in this study is lower than that of populations of the Indus Valley (14 KYA) for the STR associated diversity of R1a1a*(xM458) chromosomes calculated by Underhill et al. (2010).

Of course, these markers and other markers defining additional layers of structure under M417 have been known for over a year. Budgetary constraints and the magic of peer review combine to render this paper relatively uninformative. One of the authors explains:
I have to agree with all, but those who never tried to push an article through a serious academic journal has no idea how difficult this is. The first version was submitted like 1 year ago, and also contained pedigree rates plus 500+ FTDNA samples from different ethnic groups. But unfortunately the reviewers were so narrow-minded that we had finally to drop all FTDNA samples plus the pedigree calcs.

Personally I also do not consider Zhiv. rate valid, but I had to accept this compromise to get the paper accepted. Anyway, as Lukasz pointed out, the main goal was to introduce Z93 and Z280 into the "academic circles" so in the future we may have a comprehensive paper from a more wealthy lab. The Budapest forensics are not full of money so we had no chance to have more than 12 markers tested and "low-chance SNPs" like Z284 in Hungary. Actually we submitted the first draft before Z283 was established securely on the FTDNA tree so we could not include it later...

My comments from last year on the dna-forums postings of an Underhill(lab that brought us Zhivotovsky "evolutionary" mutation rates)-affiliated academic stand:
Another poster points out: "Dividing by 3 [to bring the estimate more in line with real mutation rates] gives an age of 3300 years, almost exactly the estimate from Nordtvedt's spreadsheet." Someone else recently estimated the TMRCA for L342.2+ at around 3,600 years. So: if current patterns hold, the bulk of South Asian R1a unambiguously falls within European R1a variation. While I fully expect, when we eventually see results for these markers in large academic samples published, the papers will feature evolutionary mutation rates and less than parsimonious attempts to fit the distribution of M417 sublineages to archaeology, it's pretty clear to me Z93 and L342.2 originated on the Steppe within the past 4000 years or so and spread with Indo-Iranian.
Again: the most straightforward interpretation of the evidence is that Z93 is a relatively young branch of an evidently European lineage. Accurate, unbiased dates using SNPs instead of STRs should be here soon enough, definitively settling this and other issues.