May 8, 2013

Telles & Ortiz: 5th generation Mexican-Americans not catching up educationally

A major empirical question about immigration is whether or not the descendants of Mexican immigrants will turn out to be prosperous people who pay in far more in taxes than they require in government services. To East Coast pundits who have only recently come into contact with large numbers of Mexicans, this seems like a slam dunk. Aren't Italian-Americans doing well? And aren't Mexicans the New Italians?

To native Southern Californians like myself, however, who grew up around a lot of Mexican-Americans way back in the 1960s, this doesn't seem like such a sure thing.

Pancho Gonzales
On the one hand, in my experience, Mexican-Americans who more or less left their culture and married into non-Hispanic families seemed to do pretty well, as my quasi-in-law Larry showed. (Let me apologize in advance for this example's genealogical complexity.) My mother's first husband was a Minnesota Swede who joined the Marines and was killed in combat on Iwo Jima. We socialized frequently with my mother's first husband's extended family, including his niece, who married in the early 1960s a Chicano named Larry and took his Spanish surname as her married name. Larry is an athletic six-footer who looks about 7/8ths white (think of Pancho Gonzalez, the great 1940-1970s tennis player from Los Angeles, and that's pretty much what Larry looks like). Starting in the 1960s, Larry worked for decades in corporate sales, selling Kraft products to supermarket headquarters. He made a solid middle class living and bought a nice home in the Santa Clarita exurb.

Or, a close friend of mine from elementary school was the son of a Central American banker who married a lady from Iceland. Another close friend had a Spanish surname too, although his dad dropped out of Yale on December 8, 1941 to enlist in the Army, so he wasn't Hispanic in anything except the Spanish surname. (The family claimed they were descended from an admiral in the Spanish Armada who was shipwrecked in Ireland in 1588.)

On the other hand, there were plenty of working class and poor Mexicans in Southern California in the 1960 who did not marry out of their cultures. Their numerous children and grandchildren have not, on the whole, set the world on fire.

The lack of high-achieving Hispanics in Southern California in 2013 is startling when you start to look for them. The current mayor is Mexican, but he's a mediocrity who took four tries to pass the bar exam. The leading candidate to replace him, the dapper Rhodes Scholar Eric Garcetti, claims to be Mexican on the grounds that an Italian ancestor got kicked out of Mexico during the Revolution one hundred years ago.

In general, what you see in California today is that while the rich are even richer than ever, there is less of the broad middle class prosperity that made California a global cultural leader in the postwar era. (Yes, I realize that non-diverse white bread Republican suburbs like Orange County couldn't, by definition, possibly have produced anything vibrant: except, they did.)

My observations have been systematically verified by a major social science project affiliated with the UCLA Chicano Studies department. The 2008 book Generations of Exclusion: Mexican Americans, Assimilation, and Race by UCLA sociologist Vilma Ortiz and Princeton sociologist Edward Telles documents a 35-year-long study of, first, a sizable 1965 sample of Mexican-Americans in Los Angeles and San Antonio, followed by a 2000 study of the 1965 sample's children, plus questions about the 2000 sample's children.

In their sample, not surprisingly, they saw rapid growth in total years of education from first generation (immigrants from Mexico) to the second generation (American-born children of the immigrants). In the 2000 sample, however, there was no further continued ascent, even unto the fifth generation. From p. 113-116 of their 2008 book:
Our data also allowed us to investigate the educational achievement of the grandchildren of the original respondents, enabling us to distinguish a fifth generation-since-immigration. ... Figure 5.3 shows that these grandchildren of the original respondents, who are third to fifth generation, seemed to be doing no better than their parents. By this time, the third, fourth, and fifth generation were performing equally, with no significant differences among them. In the third generation, 85 percent had graduated from high school, versus 84 of the fourth generation and 81 percent of the fifth generation. ...
Many have assumed that the educational inequalities are attributable to the disadvantages of poor immigrant households, but the data here show that schooling outcomes stubbornly continue at low rates even into the fourth or fifth generation. The statistics presented thus far show that the education progress of Mexican Americans does not improve over the generations. At best, given the statistical margin of error, our data show no improvement in education over the generation-since-immigration and in some cases even suggest a decline.

Here are some graphs from the book. The first demonstrates the structure of their complex study. The "Original Respondents" column represented Mexican-Americans surveyed in 1965. Back then, among the first generation (immigrants) only 30% had graduated from high school. In 1965, the children of immigrants from Mexico had a 48% rate of high school graduation, and the grandchildren a 57% rate. But, when Ortiz and Telles interviewed the children of the 1965 respondents in 2000, they found stagnation, instead. In 2000, the fourth generation Mexican-Americans in their sample of the children of their original respondents had less education than the second and third generations. And the third column shows the same stagnation for the grandchildren of the original respondents, some of whom are fifth generation Americans.

The college graduation rates makes Mexican-American lack of progress starker:

And here's a graph in Generations of Exclusion from p. 110 using a different source of data but giving similar results. These are restricted to American-born third generation or higher citizens:
The light gray bars to the left represent high school dropout rates, the shorter the better. The dark gray bars to the right represent college graduation rates, the longer the better. The dashed vertical line represents the non-Hispanic white rates. Third generation or higher Mexican-Americans in this study do slightly worse on educational attainment that African-Americans.

These findings obviously have major implications for Rector and Richwine's estimate of the fiscal effects of the Gang of Eight's amnesty bill.


SF said...

For what it's worth, Jodi Arias is hispanic. Her father is mestizo Colombian.

Red Pill Theorist said...

You see Rubin's blog today? Heritage foundation member let some serious crimethink go.

panjoomby said...

15th generation african-americants not catching up educationally.

Anonymous said...

These findings obviously have major implications for Rector and Richwine's estimate of the fiscal effects of the Gang of Eight's amnesty bill.

But haven't you heard that Richwine is a racist? He claimed that Hispanics do not close the test score gap over generations, which is an obviously racist and therefore factually false claim. So none of these facts you present matter. You could prove anything with facts.

james wilson said...

Correlating years of government school education with actual levels of education is a great mistake. Ten years was the average for Americans in 1940. Are we an improvement? Tocqueville's America, that nation which he called the best educated on earth, counted six or eight years of education--the product of literacy plus the application of work from an early age.

Mexican-American drop out rates were at one time Tocquevillian, but are now merely the product of the worst of two worlds. That's assimilation.

pat said...

You write Pancho Gonzales and Pancho Gonzalez. The actual spelling he would have used originally was Gonzàlez.

Of course he also called himself Richard not Pancho, which he considered a racial slur.

But the point is that the real Spanish spelling has an accent over the a. The reason that Gonzales became the preferred spelling is that typewriters and printers preferred fewer diacritical markings. English has the fewest such accents, probably because it has had the most books in print.

So simplified spelling is a good thing. A mark of evolution from hand writing to machine based printing.

But the computer is slowing down that evolution. A Japanese typist used to take two years to train. One of the reasons that the US has led in computers is that English being simpler was easy for early computers to handle. No longer. Their is less pressure on the Japanese and Chinese to revise their alphabets as computers become more competent. The Asian languages have placed a burden on their economies that have benefitted the US. Again no more.


pat said...

You use the phase "marry outside her culture". I think this sort of usage is part of the problem. I'm not indulging in any Korzybski like nonsense here, just a straight forward separation of genetics and other factors.

Mayor Villaraigosa I think is an Amerindian. He is Hispanic by culture in that he speaks Spanish and likes Mexican food. But even if I had done better in high school Spanish I will never be considered Hispanic. No amount of Pollo en Mole will suffice to bring me the advantages of a Hispanic classification.

OTOH Villaraigosa could identify as an Amerindian and qualify for casino profits. It's not fair.

Mario Rubio seems to me to be almost all Caucasian. Caesar Romero was also mostly Caucasian and his lover the other screen Latin Tyrone Power - was completely Caucasian. It's a mess.

But a solution is in sight. I recently took a spit test. I sent away to 23andme for a genetic screening kit. One of the things they promised was to determine my Neanderthal heritage (if any). I'm wondering if this might help the feds with their problems of differentiating between race and ethnicity.

I realize that race isn't a simple presence at a SNP of some specific polymorphism. I am not guilty of the Lewontin fallacy. But maybe some new analysis of correlated SNPs would show meaningful patterns.

When it became advantageous to be an Indian about thirty years ago the feds had to establish rules. I think you had to be one eight. In the recent Massachusetts Senatorial race Indian heritage was again a big question.

On the survey front the FBI now says there are three major races only - white, black and Asian. Hispanic is now only an ethnic category. This is the logic that decides that Villaraigosa is white even though he probably has no European ancestors.

Maybe we can get some of this on a more rational basis if we all just spit into the little tube.


Anonymous said...

With all the angelenos with a film background one would think there would be a award winning documentary on the Mexican-chicano underclass of Los angeles. A documentary which focused on the living conditions in Pacoima, boyle heights,bell and other neighborhoods blighted by illegal immigration would wake up America.

Roberto Dinero said...

The family claimed they were descended from an admiral in the Spanish Armada who was shipwrecked in Ireland in 1588.

That's a great tip for some enterprising Diego Gatsbillo--assuming anyone you meet even knows what the Spanish Armada was, they won't bother to dispute such a too-good-to-check ruse, though best not attempted if you have red hair (an established tell of deceivers)

JP Andreas, Pres. A.C.L.P. said...

Very well written & intriguing analysis which bears on the Richwine controversy, (and to a lesser extent, my own at If it holds up against any other countervailing research of the "experts," it not only appears that the pudding is well cooked, but that the folks at Heritage owe Richwine a HUGE apology. Unfortunately, the inevitable "chilling" effect on other social scientists who might tend to "self censor" in order to keep their jobs, insures that further research and "peer review" in this area and, more importantly, study of its impact on public policy, is much less likely. NOT a particularly glowing outcome for liberty, economic vitality, OR race relations in our country. Sadly, it appears short-term political advantage is more important to Heritage than the long-term economic health of the United States. jp

Bill Guerrero said...

Its racism perfect example is the San FernandoValley The Mexucan Americans are the dominate culture but who are the Congressman in the San Fernando Valley Two White guys