​

American education statistics commonly distinguish among White, Black, Hispanic, and Asian/Pacific Islander students. This is in accord with a tradition of counting the American population by race. The first U.S. Census, in 1790, divided the American population among free White males, free White females, other free persons, and slaves. Slaves were assumed to be recently imported Africans or of enslaved African descent. The Census found no slaves in Maine and Massachusetts and only 16 in Vermont. There were, however, 21,000 in New York and 11,000 in New Jersey. But most were in the southern states. There were about 100,000 each in Maryland, North and South Carolina. There were almost 300,000 in Virginia. The slave populations of the southern states approached one-third of the total in each of those states. Counting slaves was politically important because, although of course they could not vote, for purposes of congressional representation each slave counted as three-fifths of a free person, giving states like Virginia disproportionate national influence. Some things change, some stay the same. 150 years later, the authors of the 1930 Census report commented: “The distinction between white and colored is the only racial classification which has been carried through all the 15 censuses.”

In the 1930 Census “The classification of the population of the United States by color or race distinguishes in many cases only three main groups, namely, white, Negro, and ‘Other races’.” The White population in 1930 was nearly 90% of the total. There were 1.5 million Mexicans in the US in 1930 (1% of the total). There were 75,000 Chinese, 340,000 Japanese, 8,000 “Hindu” and 45,000 Filipino (the Philippines were at that time a colony of the United States). Mexicans, who were separated from the White group for the first time in 1930, were defined as “all persons born in Mexico, or having parents born in Mexico, who are not definitely white, Negro, Indian, Chinese, or Japanese.”

These historical notes inform our present day concerns. There seems to be an assumption underlying American education statistics today that the White and Black student populations are each homogenous. White students are not divided by religion or national origin, as they once were, and although Black students may be recent immigrants from Africa or the Caribbean, these differences are seldom taken into account. Hispanic and Asian/Pacific Islander students, are also grouped without further differentiation. This is particularly unsatisfactory for the Hispanic population in places like southern Florida, where residents may identify as Cuban, Columbian, Nicaraguan, Dominican, Costa Rican, or “other Latin American” of many varieties. “Asian” is an even less convincing category. While Hispanics might have the Spanish language in common (or a family history of speaking Spanish), Asian Indian-Americans and Chinese-Americans, for example, have less in common with one another than do White and Black Americans, who, at least, share a common history and language.

The varieties of Asian Americans are of interest for the analysis of educational differences as the educational achievements of Asian students are sometimes used as a challenging standard for other, especially Black and Hispanic, students. If “Asian” students can do well, why do not Black and Hispanic students do better? It must be, as the late Senator Moynihan implied, a problem with the Black family—their problem, not an issue for governmental concern. It seems often taken for granted, on the basis of what might be called “non-critical race theory,” or just racism, for short, that “Asian” students excel in school and college, perhaps, it isn’t said, because of something “cultural”. And, indeed, many students classified under that rubric do excel. But, of course, not all, and there is the troubling matter that “Asian” is something less than an exact term. It means little more than students who are neither “White,” “Black,” or “Hispanic”. Within it there are the Americans of East Asian family origin—Chinese, Korean, Japanese, for example; those of Southeast Asian family origin—Vietnamese and Hmong, to pick two; and those of South Asian family origin—chiefly what the Census refers to as Asian Indian.

In order to learn something about these distinctions and their implications, we will proceed stepwise, as it were, beginning with an analysis of National Assessment of Educational Progress (NAEP) grade eight reading assessments for “Asian” students, then focusing on Census (American Community Survey) data. Although the Census does not provide information about educational achievement, it does provide information concerning educational attainment by geography (city, state, national) and by detailed race and ethnicity, as well as many other factors.

To begin with NAEP: Students are scored by NAEP as having reached the “Advanced” level in their grade, “Proficient”, “Basic,” or below “Basic”. Students whose schools have brought them to the Proficient or Advanced levels have been called “high performers.” They are likely to graduate from high school and go on to earn Bachelor’s degrees or beyond. Students whose schools have left them below Basic have not been taught the skills required at their grade level. They may very well be handicapped by this for the rest of their lives, unable to read fluently and so unable to easily obtain skilled employment or higher education. And this for their children as well.

According to the National Assessment of Educational Progress, both urban and suburban schools teach more than two-thirds of Asian/Pacific Islander students from more prosperous families well enough for them to be classified as high performers (“Proficient” or “Advanced”), while fewer than 10% are left at the below Basic level, unable to read materials expected of middle school students. On the other hand, just one-third of Asian/Pacific Islander students whose family incomes are low enough to qualify them for the National Lunch Program are taught well enough to be assessed as high performers in both urban and suburban schools. Among Asian/Pacific Islander students, as with others, higher family income accompanies higher reading skills acquisition in middle school.

Nationally, more than half of Asian/Pacific Islander students are assessed at or above Proficient in grade eight reading, 13 percentage points higher than White students. The gap is even larger at the Advanced level: 12 percent of Asian students reaching that goal, as compared to only 5 percent of White students. Among individual states, the gap in the percentage above Proficient reaches 20 percentage points or more in four states and the gap is only reversed in four: Minnesota, with its Hmong population; Rhode Island; Utah; and Hawaii, with its history of indentured servitude of Japanese farm workers.

In order to look more closely at the “Asian” category, we must turn from NAEP to Census data. In the following, the data by national origin will be compared with the overall grouping as “Asian alone” and, as the usual standard for comparison among racial and ethnic groups is that of White, non-Hispanic, performance, that, also, will be provided, under what the Census calls “White alone”.

For reference, here are the population statistic for the largest Asian groups with “Asian alone” and “White alone” added for purposes of comparison:

National Origin Population
Asian Indian 4.2 million
Chinese 4.4 million
Filipino 3.0 million
Japanese 0.8 million
Korean 1.5 million
Vietnamese 1.9 million
“Asian alone” 18.6 million
“White alone” 236 million

Out of 18.6 million Americans counted by the Census as “Asian Alone,” slightly more than half are either Asian Indian, Chinese or Filipino.

Because educational achievement and attainment is widely believed to depend on conditions in the home, it is relevant to examine marriage, divorce and out-of-wedlock childbearing statistics. Nearly 60% of all Asian adults over the age of 15 are married, compared to just over 50% of White adults. Only 4% of Asian adults are divorced, compared to 10% of White adults. Just 10% of unmarried Asian women 15 to 50 years of age have had a child; for Whites it is 27%. White, non-Hispanic, Americans are less likely to be married than American Asian/Pacific Islanders, more likely to be divorced; more likely to have children out of wedlock.

Do all Asian/Pacific Islanders show similar statistics, or are their notable variations?

Of course there are variations.

National Origin Marriage Rate Divorce Rate Unwed Mothers
Asian Indian 68% 3% 3%
Chinese 57% 5% 7%
Filipino 55% 7% 18%
Japanese 57% 9% 8%
Korean 56% 6% 11%
Vietnamese 53% 7% 15%
Hmong 40% 4% 37%
“Asian alone” 58% 5% 10%
“White alone” 51% 12% 27%

More than two-thirds of Asian Indian adults are married, only 3% are divorced and just 3% of Asian Indian women with children are unwed mothers. The divorce rate of Japanese-Americans is three times higher and the percentage of Filipina unwed mothers is six times higher than that of Asian Indian women. It is obvious that the category “Asian alone” for these data conceals more than it reveals. Notice the comparative statistics for “White alone” adults. We might ask the successors of Senator Moynihan if there is a cultural problem with the White family.

As we have seen, family income appears to have a particularly positive effect on the educational achievement and attainment of children. The following table shows variations in family income and poverty rates among the different groups classified as “Asian.”



National Origin Median Family Income Family Poverty Rate
Asian Indian $141,100 4%
Chinese $105,600 8%
Filipino $108,700 4%
Japanese $112,300 4%
Korean $99,300 7%
Vietnamese $80,200 8%
“Asian alone” $107,200 6.5%
“White alone” $86,000 7%

All the principal Asian groups have median family incomes greater than that of “White alone” families except the Vietnamese. The incomes of Asian Indian families are much greater than “White alone” or those of any of the other Asian groups. Similarly, the family poverty rate of Asian Indians is lower than the others (although Japanese and Filipino Americans are close). Three out of four Chinese Americans and Asian Indians are foreign born, while fewer than half of Japanese Americans are immigrants. Chinese-Americans have both relatively high family incomes and relatively high family poverty rates, indicating some diversity among them, perhaps based on when their families immigrated to the United States and the resources they had with them when they did. Forty-two percent of Chinese Americans report that they speak English “less than very well,” compared to just 22% of Japanese Americans and 18% of Asian Indians.

Education, then: If the category of high performance based on the NAEP can be taken as roughly analogous to that of students who go on to a Bachelor’s degree or above, as recorded by the Census, we find that 76% of Asian Indians, but just 58% of Chinese Americans attain that standard. It is reached by 50% of Filipinos, 32% of Vietnamese, 59% of Koreans and 54% of Japanese. In other words, the distribution is all over the map, as it were.


National Origin Percent in College Percent BA & +
Asian Indian 32% 76%
Chinese 48% 58%
Filipino 42% 50%
Japanese 48% 54%
Korean 45% 59%
Vietnamese 40% 32%
“Asian alone” 40% 56%
“White alone” 27% 34%

The Vietnamese are relatively new arrivals in the United States, many as refugees from the war, and are still working their way through the American education pipeline. By way of a comparison, there is a smaller percentage of “White alone” Americans than any Asian group in college and a relatively small percentage with higher educations. This perhaps is both a cause and an effect of lower median family incomes.

Most of the children of the 18 million “Asian” Americans—5% of the population—live in wealthier, more stable, more highly educated families than White, Hispanic or Black students. The home situation for students in most of the Asian national origin groups establishes a benevolent feedback loop: prosperous, highly educated parents make it likely that their children will grow up to be prosperous and highly educated.

Asian Indian Americans are helping to drive these positive outcomes for “Asian” Americans. Most Asian Indian adults, 3 million, are foreign born. They are, then, for the most part recent arrivals in the U.S., many arriving on H-1B visas, issued to those who wish to perform services in a specialty occupation. In a recent year, Asian Indians were issued 72% of H-1B visas. This may account for the relatively small percentage in college while at the same time a very high percentage have Bachelor’s degrees or above. As a group, they are unusually accomplished. In spite of the fact 76% report that the language spoken in their homes is one other than English, 72% of Asian American adults are employed in management, business, science, etc., compared to 42% of “White alone” adults. 99% of Asian Indians live in households with a computer and 97% in homes with broadband internet subscriptions. They will have arrived in this country, college degrees already in hand, qualifying for the much sought after H-1B visa, rapidly taking their place in the technology savvy American upper middle class.

It takes nothing away from this Asian Indian record of achievement to point out that it is most likely unique, and although something for all Americans to strive for, it is problematic as a standard for comparison with students from other groups, either native or foreign-born, who do not benefit from similar advantages.

Which brings us to a similar analysis for Hispanic Americans. Unlike Asian Americans, whose national origin is distributed among the two “clusters” of Asian Indians, Chinese and Filipinos at 3 to 4 million and Japanese, Korean and Vietnamese at between one and two million, Hispanics are overwhelmingly members of one group, those of Mexican origin, only distantly followed by Puerto Ricans. For reference, here are the population statistics for the largest Hispanic groups with “Hispanic alone” and “White alone” included for purposes of comparison:

National Origin Population
Mexican 37.2 million
Puerto Rican 5.8 million
Cuban 2.4 million
Dominican 2.1 million
Guatemalan 1.7 million
Honduran 1.1 million
“Hispanic” 60.5 million
“White alone” 236 million

A third (12.6 million) of the Mexican population of the United States resides in California. Another 9.4 million are residents of Texas and there are 2 million Hispanics in Arizona, about a third of the residents of these states. These border states were once part of Mexico. If present trends continue it will not be long before those states have majority Mexican populations once again.

Now for family statistics:

National Origin Marriage Rate Divorce Rate
Mexican 43% 7.5%
Puerto Rican 35.5% 11%
Cuban 44% 13%
Dominican 39% 11%
Guatemalan 42% 5%
Honduran 42% 6%
“Hispanic” 43% 9%
“White alone” 51% 12%

These, like other Hispanic data categories, are driven by the Mexican distributions.
Mexican Americans marry at lower rates than the “White alone” group and also divorce at lower rates. The marriage rates of the other Hispanic groups are lower than those of the White group, but the divorce rates are highly variable. In general, the Central American groups have comparatively lower divorce rates than other Hispanic or White, non-Hispanic, Americans .

Looking at the “economic” topic of socio-economic analysis, we find that all the Hispanic groups have median family incomes considerably below that of “White alone” and much higher family poverty rates. This may well be a factor in their comparatively lower levels of education attainment.

National Origin Median Family Income Family Poverty Rate
Mexican $58,500 15%
Puerto Rican $58,700 17%
Cuban $66,400 11%
Dominican $51,500 18%
Guatemalan $47,800 21%
Honduran $45,800 24%
“Hispanic” $59,600 15%
“White alone” $86,000 6.8%

Finally, Hispanic education data presents a population that except for the Cuban group is less likely to be in college than the “White alone” group and considerably less likely to have a Bachelor’s degree or above (again with the exception of the Cuban group).

National Origin Percent in College Percent BA & +
Mexican 20% 13%
Puerto Rican 22% 20%
Cuban 31% 30%
Dominican 25% 20%
Guatemalan 16% 10%
Honduran 16% 12%
“Hispanic” 22% 18%
“White alone” 27% 34%

* * *

This analysis indicates that both the commonly used Asian and Hispanic student groups are actually highly diverse. Within the Asian group, the Vietnamese group has not yet reached the level of achievement of the others, while Asian Indians far exceed the average. This is an indication that generalizations about “Asians” as a model minority should be used only with caution.

Within the Hispanic group there are two special situations. Nationally, Central Americans have not yet reached the level of achievement of some others, such as the Cubans, perhaps reflecting the differing condition of educational institutions in their home countries. The concentration of Mexican Americans in California and, to a lesser extent, Arizona and Texas, calls for specific attention to their educational needs, as educators in those states need no reminding.

Of course, none of this tells us anything about individual students. However, it may help us avoid mistaken assumptions arising from broad-stroke clichés about race and ethnicity in our schools and inadvisability of invidious comparisons. It is pointless to talk about “Asians.” It is one of the concepts, fossilized racism, that bedevils this country. The useful categories—useful for the students themselves—are socio-economic: the level of parental education and their economic resources.

Chinese American students whose parents are highly educated and prosperous have similar educational needs to those of Asian Indian students whose parents are highly educated and prosperous, to those of Hispanic, White and Black students whose parents are highly educated and prosperous. Schools, as they are, are successful in supporting them as high performers, on the way to college and postgraduate education, successful careers and the probability of passing these resources to the next generation. Chinese American students whose parents are not highly educated and prosperous have similar educational needs to those of Asian Indian students whose parents are not highly educated and prosperous, to those of Hispanic, White and Black students whose parents are not highly educated and prosperous. Schools, as they are today, are unsuccessful in supporting them to be high performers, on the way to college and postgraduate education, successful careers and the probability of passing these resources to the next generation. Our schools can fulfill their missions by providing well-known additional resources for these students: small classes, extended school days, weeks and years, teachers with specific training to meet their needs.

Michael Holzman