Wednesday, 3 September 2014

The United States of Mexico

Have the lower classes been at it again, breeding like rabbits, to the detriment of the gene pool? Have the upper classes, far from defending their own fold, been inviting in hordes of lower class foreigners, to breed like even more fecund rabbits, thus lowering the gene pool even further, and in the process undercutting the wages of their local proletariat, to the benefit of the rentier class? According to Irwin Kirsch, Henry Braun, and Kentaro Yamamoto, and Andrew Sum something of that sort may be happening in the USA, that Western outpost of Europe set up by English non-conformists and free thinkers. The authors have written a policy report for the Education Testing Service with the uplifting title America’s Perfect Storm: Three Forces Changing Our Nation’s Future.

The three forces they identify are: race gaps in educational achievement; the high premium paid to the well-educated; and immigration.

The policy paper is well written, clear, and cautious in what it says, despite the dramatic title. Their language is absolutely nothing like my introduction to their paper, because in line with contemporary sensibilities it is written in a vacuum as regards intelligence and educability. For example, they describe the three threats as: divergent skill distributions, the changing economy, and demographic trends. However, when you look at their data they are talking about intelligence and racial differences, the premium on intelligence, and the inflow of Mexicans to the USA. Stripped of circumlocution, they are perturbed by African Americans and Mexicans who are going to create problems which are difficult to solve by conventional means. Nonetheless, they say the accepted right things, which is that steps must be taken to make the convoy of different ability tribes stick together. Far from being a total melting pot, Americans tend to spread out a bit, putting some distance between disparate groups, something the wide expanse of the continent facilitates.

Here are the author’s words, drawn from different sections of their report:

Given our country’s growing demographic diversity, [we could] imagine our nation as a convoy. Some of the boats are large, well built, and able to ride out the heaviest of seas. Others are somewhat smaller, less well-equipped, but still quite sturdy. But many are fragile, meagerly equipped, and easily swamped in rough waters. That convoy — the individuals, families and communities that make up our nation — is in the midst of a “perfect storm,” the result of the confluence of three powerful sets of forces: divergent skill distributions, a changing economy, and demographic trends.

Data from the National Assessment of Educational Progress (NAEP) reveal that between 1984 and 2004 reading scores among 13- and 17-year-olds remained flat, and the achievement gaps were large and relatively stable. For mathematics the story is only slightly different. While the mean scores for both the nation’s 13- and 17-year-olds improved slightly, they did so across all groups, with the result that the average size of the Black-White and Hispanic-White achievement gaps remained large and relatively stable.

Comment: Heiner Rindermann and I had a look at these scores, and found that although scores did improve in the early 80s (for no very clear reason) the differences between Hispanics/African Americas and Whites/Asians remains large.

National surveys of our adult population indicate that large numbers of our nation’s adults, 16 years of age and older, do not demonstrate sufficient literacy and numeracy skills needed to fully participate in an increasingly competitive work environment. These skills are also needed to function effectively in our complex society, with its large bureaucratic institutions and its complex legal, health care, and retirement systems.

• More importantly, these skills are not evenly distributed across groups defined by race/ethnicity, country of birth, and socioeconomic status. In fact, there are substantial differences in average proficiencies among these groups that influence their social, educational, and economic opportunities.

• International surveys of student and adult populations indicate that while our average performance is no better than mediocre, our degree of inequality (the gap between our best and least proficient) is among the highest in OECD countries.

The second force comprises the seismic changes in our economy that have resulted in new sources of wealth, novel patterns of international trade, and a shift in the balance between capital and labor. These changes have been driven by both technological innovation and globalization, resulting in a profound restructuring of the U.S. workplace. Indeed, the labor markets of today are markedly different from those of earlier decades. For example:

• In 1950, manufacturing’s share of total employment in the United States was 33.1 percent. By 1989, it was down to 18.2 percent and, by 2003, it was 10.7 percent.

• Between 1984 and 2000 the number of employed persons 16 years of age and older grew by 29 percent, or some 30 million. At the same time, employment in jobs associated with college-level education grew by some 20 million, accounting for two-thirds of the job growth.

• The country’s employment growth is expected to continue through the rest of this decade and into the next, with college labor market clusters (professional, management, technical, and high level sales) expected to generate about 46 percent of all job growth between 2004 and 2014.

Fuelled both by higher birth rates and by immigration, the Hispanic share of the population is expected to grow from 14 percent in 2005 to slightly more than 20 percent by 2030.

In 2004, nearly 57 percent of the 16- to 64-year-old Hispanic population in the United States was foreign-born, up from 46 percent in 1990. More than half of these immigrant Hispanics lacked a high school diploma.

• The lack of a high school diploma by such a large proportion of Hispanic immigrants is of concern given the fact that almost 80 percent of immigrants who have not earned a high school diploma report not speaking English well or at all.

Comment: Mexicans are unlikely to bound ahead scholastically if the last 6 generations are any guide to the future. Greg Cochran has an interesting post on this, with a long term sample which at least measures college level qualifications. This is in line with other findings on Hispanic progress.

The authors show the result for “literacy”, for which the better description would probably be everyday intellectual ability. Levels 4 and 5 have been merged because of their rarity, but the white figures trace the familiar bell curve even on four categories.


Those in the 4/5 high ability level: White 17%, Asian 9%, Black 3%, Hispanic 3% give you a relative indication of what proportion of each race will flourish in the global economy. If you want to be gloomy, look at those in the low 1 and 2 categories, which are those who will flounder in any modern economy: Hispanic 82%, Black 77%, Asians 61%. (Asian scores seem a bit low in this survey).

Having shown these results, the authors then forget about them, and in Tables 1 and 2 talk about the relatively low standing of the US compared with other nations on the PISA scores. This is all very well at the national level. However, the racial breakdown of those scores shows that White kids are doing pretty well: the national scores are dragged down by Blacks and Hispanics.

The authors then project what the figures might be like in 2030 if current trends continue. That is a big “if”, but illustrative nonetheless.


There will be significant losses of ability at the higher levels. The US will become dumber.


Figure 5 shows the figures for out of wedlock births, often seen as a proxy for absent fathers. The economist Thomas Sowell point out that these figures were much lower for Black families decades ago. However, the current picture is unlikely to boost scholastic

I am in sympathy with the points that the authors have made, and agree that they are raising important issues. I also think that this paper has been written in hostage language. The captive is led before the cameras and with unconvincing passions sings the praises of his captors, praises their humane treatment, heaps criticism on unrelated parties, usually focussing on his home government, and then get decapitated. He cannot talk straight because he fears, quite rightly, the horrors which threaten him.

These authors know full well that the educational system is not going to be able to pull a rabbit out of a hat. That hope was a bit of a stretch even in the early 70s when funding for compensatory education increased significantly. The 80’s showed some improvements, for unclear reasons, but after that very little gains were made. It seems that such low hanging fruit as was available has been picked already. Education cannot compensate for differences in ability. Rather as a hostage blinks out the Morse code message “Help Me” the authors are begging for change (and I think they don’t mean cunning new ways of getting rap artists to teach English Literature).

Unless we are willing to make substantial changes, the next generation of Americans, on average, will be less literate and have a harder time sustaining existing standards of living.

These are stirring words, but the authors do not spell out the substantial changes which would be required. These might include screening would-be immigrants for educational achievements and intellectual abilities. They might include giving up racially-based compensatory education policies and putting those funds into the general education budget, with the stipulation that the amount of public money spent on each child should be roughly the same. They might include strengthening the examination systems to ensure absolute rigour, and entry to college based on common, exam-based criteria. Perhaps some of you can provide some more detailed policy suggestions.


I seem to remember that in 2013 some guy said that Hispanics were not achieving the scholastic standards of white Americans, and was drummed out of his job.

As far as I can see, these three authors are raising the same issue, but have wisely, for their own preservation, done so using “demographic diversity” as a code for race, and “literacy” as a code for intelligence. I hope they will not be incommoded by the uncouth rabble.

Can a society which cannot bring itself to mention intelligence continue to be an intelligent society?


  1. The effect of these demographic changes may be exacerbated if, as seems probable, technological changes such as robotics continue to reduce the demand for low-skill labor.

    On the other hand if the peak-oil alarmists are correct maybe we will have to abandon an energy-intensive agriculture and go back to a labor-intensive agriculture thus soking up a lot of unskilled labor.

    When I was a child lots of people were still employed pumping gas or operating elevators. The MacDonald's of the future may be staffed mostly by robots.

    1. Robotics will probably make a difference, but generally new jobs of some sort are created. I used to worry about peak oil, but I don't think a likely scenario for this century. There may be temporary shortages, but there seem to be lots of hydrocarbons relatively near the surface.

  2. It's important too to consider the racial composition of Mexicans.

    According to the CIA World Fact Book, Mexico is:

    60% mestizo
    30% Amerindian
    Less than 10% European (mostly Spaniard)

    And what is the ancestry of mestizos? Examining genetic ancestral markers, Rubén Lisker found the average admixture of a lower-income mestizos in Mexico City to be:

    59% Amerindian
    34% European
    and 6% black

    1. I hadn't got into the genetics, but a detailed genetic picture would certainly be interesting, and the big prize will be to link the genome with behaviour, including intelligence. This cannot be done with any precision at the moment. Only 3% of the variance can be accounted for when associations derived from a sample of discovery are applied to another test population.

  3. I assume you are aware of the controversy a couple of years ago over Ron Unz’s criticism of Lynn’s ideas and about Mexican intelligence in general. Did you write anything on this?

  4. The Ron Unz critique is about the variability of National mean scores in Europe, but I think that looking at the larger samples there is a reasonable concordance, though there is a tendency for some regional differences in Europe as a whole, and also within nation differences. North of Italy seems more intelligent than the south, perhaps, Lynn argues, because of admixture with North Africa. In all these studies one has to look at the representativeness and size of the samples. More recent genetic studies of Europe show that regional differences can be tracked in detail, so it may be possible, eventually, to match the genetics to the underlying g scores.
    As to Mexico, all the above strictures apply, but the Lynn Backoff and Contreras (2005) paper suggests 94 for Mestizos, 83 for Indios, so it is the actual genetic mix of the immigrants to the US which might be relevant here.

  5. I don't think my comment posted so I'll rewrite it.

    Surely the poor literary proficiency level of hispanics has a lot to do with them not speaking English as a first language.

  6. Literacy is a misnomer, so as to avoid mentioning intelligence. The tasks are basic problem solving tests, with minimal language content. On other grounds, (paper which will be presented at ISIR 2014) it seems that immigrants everywhere are not lower in verbal than non-verbal skills, so it seems to be general ability which is lower. Equally, even if one allows 5 years to learn language skills partly required in face to face testing, one gets similar results.

    1. In that case I'm a little surprised since I believed hispanics to be half way between blacks and whites in IQ. In that study, the blacks have less in 1 and more in 3 and therefore a higher average, right? This doesn't square with all the other research findings on black and hispanic IQ.

    2. well, not necessarily a higher average but in this case, yes.

  7. And you have to wonder what kind of Asians they are. They can't be just NE Asians.

  8. The ETS report was written in 2007. Your blog gives the impression that it is recent and that the public reaction to it is unknown.

  9. Dear Quant, You are absolutely right, and all I can say now is that it brings us a mere 16 years till we can evaluate their prediction about 2030. In fact, NAEP data or similar scholastic results over the next few years should provide an indication of the trend. Other data from PISA shows persistent scholastic gaps for immigrant groups after 2 generations, which argues in favour of the authors' concerns. That is, gaps vis a vis the locals generally close in the second generation, though not in the US, Will present on this in December at ISIR 2014, most probably.

  10. Dear James,

    World has been improving all the time, and dramatically improving the last 200 years, with the same problems we face now. And we have more weapons to fight now than ever.

    The question is, are people all over the world improving their literacy? I think so. Are people from some developed country worsened their literacy? Maybe. But we are talking about averages. Could we imagine a continent like Africa improving its literacy to let the people start new studies even new business.

    Anyway is a very interesting paper to think about it.

    Regards from Lima.