Stanford University

News Service



CONTACT: Kathleen O'Toole, News Service (650) 725-1939;

Demographer explores instability of American ethnic identities

"Many assume that American Indians are a 'vanishing race,' but after 500 years of steady decline, American Indians are now more numerous than perhaps at any time in this nation's history," Stanford sociologist C. Matthew Snipp told President Clinton's panel on race relations when it met in San Jose earlier this month. The panel was formed to open a broader national dialogue on race and Snipp was among those asked to speak.

A demographer who joined Stanford's sociology department in 1996, Snipp is among a small group of scholars who have demonstrated how and why Americans' racial or ethnic identities are becoming "more complex, changeable and subject to interest-based mobilization," says colleague Cecilia Ridgeway, who chaired the department when Snipp was recruited from the University of Wisconsin. An expert on the sociology of American Indians, Snipp has shown that urban migration has contributed to a substantial increase in the number of American Indians, partly as a result of more marriages across traditional racial categories.

Because of increasing rates of intermarriage of all racial and ethnic groups in the United States, "racial divisions once considered fixed and immutable are becoming fluid and unstable," Snipp wrote in a recent journal article. Such changes become a challenge for public policies such as affirmative action that use racial identity to allocate resources or opportunities.

Visual evidence of what he has found in more formal studies is obvious on many college campuses today, Snipp said recently, as he nodded toward his office window on the Quad. "All you have to do is walk across the Stanford campus and you see these kids who look vaguely Hispanic or faintly Asian. They are racially very ambiguous in terms of outward appearances because one parent is black or Hispanic, another parent is Asian or white. We have all these combinations and so when these kids come to identify themselves, it's a hard choice, particularly when you have a federal government that has said you've got one race in your life, pick one and that's what you're going to report."

The recent history of American Indians, he said, offers a clue to the future for other groups as well. An overlooked dimension of our heterogeneous society "involves persons who may toggle between several identities as the need arises."

Thousands of Americans have apparently done just that in the years since 1960, when the United States began asking citizens to identify their own race in the decennial census. Between 1970 and 1980, the census recorded a 72 percent increase in the number of American Indians, for example. The population is still growing so that today there are more than 2 million American Indians and Alaska Natives, half of whom live in multi-ethnic urban areas, and more than half of the married adults among them are married to someone who does not identify him or herself as an Indian, Snipp said. As a result, a group once thought to be nearing extinction is now larger than it has been for several centuries.

"The census bureau's original take on this was that the data were wrong ­ that people who were changing their identity weren't real Indians," Snipp says about discussions he had with census officials in the early 1980s. Demographers eventually attributed 28 percent of the 1980 census increase in American Indians to natural phenomena ­ higher birth rates and longer life spans, he said, and they also uncovered some misunderstandings that led to errors.

"There was a large increase of Indians, for example, in a suburb of New Jersey. It turned out they were Asian Indians. They had lived in America long enough to be American citizens and so now, they thought, they were American Indians."

But an increase of 600,000 people from one census to the next was too large to attribute to widespread misunderstandings. At the time Snipp made a point to government statisticians: "In order to say these people aren't real Indians, you have to establish what a real Indian is, and that is a problem in its own right."

In other words, Snipp said, the government should not assume that people make "errors" on the census or other types of forms and surveys that ask about their ethnic identity, but rather, that they are submitting "corrections."

Snipp is an enrolled member of the Oklahoma Cherokee nation, even though he grew up in California's San Joaquin Valley. Like many Oklahomans after World War II, his father sought employment in California and ended up working in the oil fields near Bakersfield. There was also a concerted effort by the government in the 1950s and 1960s to relocate many Indians from economically poor rural Indian reservations to cities, where it was assumed they would do better economically. Those relocations have not greatly reduced the income inequality of Indians as a whole, and urban Indians generally still regard the rural "Indian country" from which they or their ancestors came as their home, Snipp told the president's panel on race. Many of the models developed to address racial inequality issues, he said, do not hold up well when tested in the context of the American Indian population.

Changing racial stigmas and preferences, Snipp believes, have affected how people identify themselves. For example, while his own parents made efforts to expose him to his Indian heritage when he was growing up in the 1950s, many older Indians experienced a childhood of being punished in boarding schools for speaking their own language. The 1960s and '70s saw a resurgence of ethnic pride and a "pan-Indian" movement in the United States that probably prompted many in several age groups to be more public about their Indian identities and to re-establish ties.

In Oklahoma, the state with the largest Indian population and where his parents took him and his brother to visit frequently when they were boys, it was not uncommon, he said, to meet Indians older than himself who thought of themselves as "a darker version of white," because of the nation's preoccupation with drawing distinctions between blacks and whites. "Oklahoma was a Jim Crow state when it was established in 1907 and the Oklahoma constitution mandated, for all intents and purposes, that Indians could be counted as white people," he said. "Where there were segregated facilities, you didn't have to say they were black. They could sit in the white section of the movie theater."

Data about multiracial ancestry is scarce, Snipp says, but census data indicate that interracial marriages among all Americans rose from less than 1 percent in 1970 to over 7 percent in 1990. First-generation immigrants often have spouses when they arrive or connections to marriage markets in their home country, but intermarriage increases with later generations, Snipp said. Large cities bring together a diverse collection of potential marriage partners also, thus making intermarriage more likely as the nation becomes less rural. About a quarter of coupled Hispanics were married to non-Hispanics by 1992 and more than half of American Indians are in such relationships. Intermarriage rates have been lowest for African Americans, but that too is changing, he said. By 1989, about 13 percent of all children born to non-whites in the United States were multiracial.

Snipp himself is not married but his brother is married to a Japanese American, he said, and his mother was wondering aloud at Christmas time about how her 2-year-old granddaughter would come to understand her identity. It is an issue that has prompted organizations of multiracial Americans and their parents to demand Congress permit multiracial identifications in the 2000 census. Recently the Office of Management and Budget issued a directive changing the instructions for the census and other government forms. The government will now ask people to check all races that apply. Snipp, who is a member of an advisory board on racial enumeration to the census bureau, said the government has yet to decide how it will tally the results.

"The bureau thinks the change probably won't impact their counts very much, and they are probably right," he said. Preliminary sampling by the bureau indicated the impact would be much larger if the government had decided to include a category called "multi-racial." In the sample surveys, he said, some people did change their identities from one time to the next, depending upon how the question was asked. He hopes to do more analysis of the data to learn about how people conceive of their ethnic identities.

At a December meeting of the advisory board, he said, the question came up about how to tally those who will now say they are several races. Snipp explained the problem this way: "Are you going to have universities like Stanford who award five fellowships to five students [who check off two races each] report giving 10 fellowships?" Or , he laughs, "will we have .5 students walking around?"

There have been allegations of ethnic fraud in states where universities give some preferences to underrepresented minorities. The Detroit Free Press, for instance, published stories in 1992 alleging fraud by several dozen University of Michigan students who claimed to be American Indian, perhaps only to get free tuition under the university's policy. While some Indians have argued that employers, educational institutions and government agencies should verify claims of being Indian, Snipp sees that as a Pandora's box.

"Self-identification is a problem for disadvantaged groups, in the sense that you have people who fraudulently claim to be part of the group," he says. "It works to the benefit of the federal government and employers because if absolves them of the responsibility of verifying these claims. But trying to adjudicate or arbitrate a claim that someone is more entitled than someone else would be a nightmare."

The pitfalls of trying to establish a pseudo-scientific way of measuring ethnicity is something American Indians have already faced because of treaties with the U.S. government. The Bureau of Indian Affairs has an elaborate structure for determining eligibility for tribal benefits on the basis of tribal ancestry, Snipp says, but it is fraught with problems and controversies. "I guess my own take is that you allow people to identify and you hope people will identify truthfully in terms of adopting a category that is meaningful to them."

Given all the problems, some commentators who advocate a "color-blind" society have urged the government to drop attempts to enumerate races or ethnic groups. Snipp points out that civil rights groups requested ending racial designations in the 1970 census. "At the last minute, they realized that if you didn't ask questions about race, you would never be able to show that black people have more unemployment, worse health, lower incomes and on and on."

The counts are important, he said, because they have "real, live social implications," including how we understand the risks associated with cancer and other diseases. "Racial differences exist, and they don't exist because we happen to ask questions about them," he says firmly.

But is affirmative action and other race-based policy still possible if race, as a social concept, is so unstable?

"I don't think the difficulties of this is a justification for dismantling programs like affirmative action," he said. "They exist because of the importance of race and the harm that has been done as a consequence of segregation, or what you might think of as the American system of apartheid that existed for 150 years."

There are no simple prescriptions, he says. His own is "vigilance, sensitivity and a deeper understanding of racial and ethnic relations." And, he added in the recent journal article, a little "luck" wouldn't hurt either.


By Kathleen O'Toole

© Stanford University. All Rights Reserved. Stanford, CA 94305. (650) 723-2300. Terms of Use  |  Copyright Complaints