May 7, 2003
Press Release
Washington, DC - Today, Congresswoman Carolyn B. Maloney (D-NY), former Ranking Member of the Census Subcommittee, praised the Census Bureau's decision to finally release estimates on the Hispanic subgroup based on the Census 2000 long form data. Because the Census Bureau changed the Hispanic origin question from the 1990 census to the 2000 census without first testing it, the number of persons listing a subgroup (such as "Mexican," Salvadoran," and "Dominican") went down dramatically. Representatives Maloney, Serrano (D-NY), Rangel (D-NY), Waxman (D-CA), Davis (D-IL), and Gonzales (D-TX) urged Director Kincannon to release the information.

"This is what happens when the Census Bureau doesn't test their changes. Now that they have finally done this report, providing better data on Hispanic subgroups, we need to make sure that in 2010, the Census bureau asks tested questions, that allow members of Hispanic subgroups to be counted accurately," said Congresswoman Carolyn B. Maloney in a statement today.

Congressman José Serrano added, "This is not a statistic that we should take lightly. Incorrect Census data has serious implications for the entire state of New York and Hispanic communities throughout the nation. We must all work together to give the Census bureau the tools to correct these errors so that they don't happen again. We will continue to work with the Census Bureau to find more accurate methods to count all Hispanics in this country. This is a very important process, one that has lasting implications for communities across the country, so we must continue to prioritize improving the Census."

Congressman Charles B. Rangel said, "The Census Bureau estimates are welcome because they provide us with needed information on the varied identities of our rapidly growing Hispanic community. The Census Bureau must now ensure its capacity to obtain accurate information on the specific national origin of each of the persons of Hispanic origin who designate themselves as such so that we as a nation can better recognize, understand and appreciate the diversity of our Hispanic neighbors."

Read the Census Bureau report

Directory of Related Information

The Census Bureau changed the Hispanic origin question from the 1990 census to the 2000 census. In 1990, the questionnaire listed specific Hispanic subgroups as examples for the "other" option, where the respondent could fill in a blank line. In 2000, the questionnaire listed "Yes, other Spanish/Hispanic/Latino - print group." Consequently, the number of persons listing a specific subgroup went down, and the number responding Spanish, Hispanic, and Latino went up. This was particularly troublesome for groups like Nicaraguans and Panamanians which showed a decline between 1990 and 2000 despite the large growth in Hispanics between 1990 and 2000. Conversely, the number of "other Hispanic or Latino" increased 223% between 1990 and 2000.

In response to questions from Congress and Hispanic members of the Race and Ethnic Advisory Committee, the Census Bureau undertook a project to develop better estimates of Hispanic subgroups in the 2000 census using long form information on place of birth and ancestry. These items were used to identify subgroup membership, where possible, for persons who provided the general response such as Spanish, Hispanic, or Latino.

Unfortunately, the report accompanying these estimates does not present the short form counts of Hispanics from 1990 and 2000 from which these concerns arose. Instead, the report provides estimates from the long form. These numbers differ somewhat from the short for counts. For example, the short form count for Panamanians shows a decline from 1990 to 2000, while the long form estimates shows an increase. Short form counts for Hispanic subgroups can be found in Appendix A of the Pew Hispanic Center study "Counting the 'Other Hispanics': How Many Columbians, Dominicans, Ecuadorians, Guatemalans and Salvadorans Are There in the United States" by Roberto Suro (May 9, 2002).