Shifting Identities

By Published on .

When asked their ancestry or ethnic origin, more people than ever before chose “United States� or “America� on their Census 2000 forms. Some 20 million people identify themselves with the United States or America, up 51 percent from 13 million in 1990, according to the Census 2000 Supplementary Survey.

However, that doesn't mean we're shedding our respective cultural identities. Social scientists recommend caution when interpreting ancestry data because people who select the same ancestry may differ markedly in terms of age, gender and levels of education and assimilation. The difficulty of answering a question on ancestry in a country dominated by immigrants, and the methodology used to gather the data can pose problems. For example, there's no guarantee that a person who chooses “American� as an answer to a question on ethnicity won't make an entirely different choice on another survey. “We're treating ancestry as if it were like age or sex, where there's one correct answer,� says Calvin Goldscheider, professor of sociology at Brown University, in Providence, R.I. “But increasingly, ancestry is variable. It changes.� Because we have multiple identities, it becomes difficult to pin down a person's origins, he explains. A person can say he's an American, but when he visits his Italian American grandparents, he'll say he's Italian. And what's true for young adults may change as they get older and want to share their family heritage with their children, for example.

“There is an optional character to some of these ethnic identities,� agrees Jeffrey Passel, a demographer at the Urban Institute in Washington, D.C. Because of high rates of intermarriage and distance from their immigrant roots, people increasingly have multiple ethnic identities to choose from in defining their ancestry. How they choose to characterize themselves may depend on variables such as time and place, or even the power of suggestion.

For instance, demographers know that the way an ancestry question is asked on the census form can dramatically impact results. In the case of the census, examples provided as a tool to help answer the question skewed the response, says Passel. In 2000, American or United States was not an example, but the category may have benefited from the omission of older European ancestry groups, such as German and English, from the list of examples provided that year.

Groups listed in the example in a given year often experience an increase in numbers claiming that ancestry. In 1990, for instance, German and Croatian were among the examples accompanying the ancestry question, which asks: What is your ancestry or ethnic origin? The number of Germans appeared to leap from 49 million in 1980 to 58 million in 1990, when it was included in the example, only to fall again in 2000, to 46 million, when it was left out. Something similar happened to people with roots in Croatia. The number of people of Croatian ancestry swelled 115 percent, from 252,970 in 1980 to 544,270 in 1990, when it was listed among the examples. In 2000, when it wasn't among the examples, it dipped 28 percent to 390,080. Similarly, in 1980, the first year ancestry data was collected, English was included in the sample. Since then it's been omitted and the number claiming any English ancestry has dropped precipitously from 50 million in 1980 to 28 million in 2000.

However, the numbers don't necessarily reflect real shifts in populations. “It's highly likely many either migrated into the American category or simply didn't answer the question,� says Richard Alba, professor of sociology at the State University of New York at Albany, referring to the decline in people claiming English and German roots.

Census Bureau researchers recognize that the examples they use in the question on ancestry affect response, says Angela Brittingham, demographic analyst at the bureau. Beyond rotating some of the prompts, she says, the bureau has no plans to change how the ancestry question is administered.


Italian ancestry was the first example provided on the Census 2000 questionnaire, which may have increased the response rate of those claiming that ancestry.

United States or American 19,643,045 13,039,560 6,603,485 51%
European 1,912,050 466,718 1,445,332 310%
Italian 15,903,962 14,664,550 1,239,412 8%
African 975,632 245,845 729,787 297%
Norwegian 4,547,291 3,869,395 677,896 18%
Haitian 650,207 289,521 360,686 125%
Jamaican 681,367 435,024 246,343 57%
Portuguese 1,311,008 1,153,351 157,657 14%
Brazilian 205,019 65,875 139,144 211%
Finnish 797,169 658,870 138,299 21%
Arab/Arabic 261,529 127,364 134,165 105%
Ukrainian 862,416 740,803 121,613 16%
Yugoslavian 378,333 257,994 120,339 47%
Northern European 184,628 65,993 118,635 180%
Iranian 348,741 235,521 113,220 48%
Czech 1,395,867 1,296,411 99,456 8%
Source: Census 2000 Supplementary Survey

Forecast is American Demographics' bimonthly newsletter devoted to census coverage. For a subscription, call 212-204-3877 or send an e-mail to [email protected].

Most Popular