The 2000 U.S. Population Census, conducted between January and September 2000, was, according to the U.S. Census Bureau of the Department of Commerce, “the largest peacetime effort in the history of the United States” (U.S. Census Bureau 2001). The first bits of data – overall estimates of the resident population on April 1, 2000 – were released in December 2000 (U.S. Department of Commerce 2000). Since then, the Census Bureau has released several data sets that allow for analysis of the 2000 U.S. population by a variety of demographic and socioeconomic criteria. In particular, the recent release of some of the Public Use Microdata Sample (PUMS) files permits analysis of the demographic and socioeconomic characteristics of the Cuban- American population in the United States in 2000. This paper barely skims the surface and presents a preliminary look at some of the tabulations from the PUMS data. The author hopes that the paper will generate interest by others in analyzing the wealth of information on Cuban-Americans available from the 2000 Population Census.
THE U.S. 2000 POPULATION CENSUS: DATA AVAILABILITY
The Census Bureau used two different data collection instruments, a short form administered to all households, and a long form, administered to one in six households. The short form was focused on enumeration, although it did capture basic age, sex, race and ethnicity data. The longer questionnaire included all of the questions on the short form, plus detailed questions that expanded on the topics covered by the short form and covered other areas of interest. 2 The information collected in the long form is of great interest to researchers interested in the Cuban- American population of the United States, since it includes questions about income, employment, housing, year of arrival to the United States, country of origin, and English-language ability, among other areas. The Cuban-American population comprises both persons born in Cuba who migrated to the United States (immigrants) as well as those persons who were born in the United States of Cuban descent. Information is also available for each of these two subgroups.
Data released from the short and long forms was made available throughout 2002 and 2003 in publications and releases of the Census Bureau, and some of it still awaits release. All data released until recently, however, was based on geographic boundaries. This type of data allows for analysis of the characteristics of a given state, county, city, census tract, block group, and block, but only on the basis of a geographic unit. This data may be useful to city planners and community leaders. For example, it would enable a city or county to determine how many Cuban- Americans live within its borders and also how many immigrants have arrived since 1990. But it does not allow a researcher to determine how many Cuban- Americans have arrived since 1990, or since 1959, or any other year. Nor does it permit comparisons, for example, of auto ownership between Mexican-Americans and Cuban-Americans in a given city, even if the overall number of car owners and of Mexican- Americans and Cuban-Americans in that city are available.
These limitations inherent in the geographic area data are addressed by Census’s release of PUMS, also known as the one- and five-percent samples. These data releases provide researchers with actual responses to census questionnaires that allow for customized cross-tabulations. Due to the specificity of the data, these releases widen the geographic areas to maintain confidentiality and avoid identifying any respondent. While the initial geography-based releases discussed above divide the country into 8.5 million blocks, with an average of 33 persons in each,3 the one-percent sample provides information for population concentrations of 400,000 or more only. That is, over an area with population of 400,000 or more, the one-percent sample allows a comparison of Mexican- American and Cuban-American car-owners and many other comparisons. The one-percent national sample was released between April and June 2003, and is now available to researchers.
The rest of this paper presents some tabulations and some observations of the one-percent PUMS data as it pertains to Cuban-Americans. The tables update some of the tabulations regarding the demographic and socioeconomic characteristics of the Cuban- American population carried out by academic researchers (e.g., Pérez 1985 and 1986a; Pedraza 1996) based on the 1980 and 1990 U.S. Population Census.
Table 1 presents summary demographic and socioeconomic characteristics of the Cuban-American population in the United States in 2000 compared to 1990. This tabulation includes both immigrants born in Cuba as well as persons born in the United States of Cuban descent.
• The number of Cubans in the United States in 2000 (1,254,439) was 19.1 percent higher than in 1990 (1,053,197).
• The 2000 Cuban-American population was older (median age 40.0 years v. 38.9 years in 1990), had higher educational attainment (36.9 percent of the population had less than a high school education in 2000, compared to 43.4 percent in 1990), had a lower labor force participation rate (55.5 percent in 2000 compared to 65.0 percent in 1990) and had a higher employment rate (93.9 percent compared to 92.5 percent).
• The number of Cuban households in 2000 was 474,258, 20.9 percent higher than in 1990.
• The income of Cuban households generally rose during the 1990s, with the share of households earning under $25,000 per annum declining from 45.4 percent in 1990 to 36.7 percent in 2000, while at the other extreme, the share of households earning more than $50,000 per annum rose from 23.1 percent in 1990 to 38.0 percent in 2000. • Nevertheless, the share of Cuban families living below the poverty line rose in 2000 to 14.3 percent compared to 11.4 percent in 1990.
Table 2 presents selected demographic and socioeconomic characteristics of the Cuban-American population in the United States in 2000 distinguishing between those persons born in the United States and those who were born in Cuba and migrated to the United States.
• The U.S.-born Cuban population of the United States totaled 390,705 persons in 2000 (31.1 percent) while the immigrant population totaled 863,734 (68.9 percent).
• The U.S.-born Cuban population is much younger than the immigrant cohort, has a higher labor force participation rate, has a lower unemployment rate and has attained higher educational achievement.
• A lower share of the U.S.-born population lived in families under the poverty line compared to the immigrant cohort (12.6 percent v. 15.4 percent) and also a much higher share of U.S.-born group lived in families earning five times the poverty line income or higher (28.3 percent v. 19.9 percent).
• Finally, the U.S.-born population had significantly higher representation than the immigrant population in professional and office and administrative support occupations (white collar occupations) compared to the immigrant cohort.
Table 3 presents information on selected demographic and socioeconomic characteristics of the Cuban- American population in comparison with Mexican- Americans, Puerto Ricans, All Other Hispanics, and the U.S. population at large. Based on this table, it can be concluded that in 2000:
• Cuban-Americans were substantially older than the other Hispanic cohorts and the U.S. population at large.
• Cuban-Americans had a lower labor force participation rate than the other Hispanic groups and the U.S. population at large.
• Cuban-Americans had lower unemployment rates than other Hispanic groups, but higher than the U.S. population at large.
• Cuban-Americans had higher levels of educational achievement that other Hispanic groups other than Puerto Ricans; the level of educational achievement of Cuban-Americans was below that of the U.S. population at large.
• A far lower percentage of Cuban-Americans lived in families earning below the poverty line than other Hispanic groups, but this represented a higher share than for the population at large.
• The same seems to be the case for income. Income of Cuban-Americans in 1999 was higher than for other Hispanic groups, but lower than for the U.S. population at large.
This paper is a first effort to produce some tabulations from the U.S. 2000 Population Census for the Cuban-American population. We have made some observations regarding the characteristics of the Cuban- American population from comparisons with the 1990 Census and with other ethnic cohorts. Subsequent investigations should include, inter alia, isolating the different waves of Cuban migrants. In particular, developing additional tabulations for a range of indicators broken down by period of entry of Cuban immigrants would be extremely useful. The release of the PUMS five-percent sample, which is scheduled for the fall of 2003, will provide another rich source of data on the Cuban-American population, and will allow for even more precise national level analysis. Another area of potential research is comparisons of the characteristics of the Cuban- American population in Florida and South Florida with Cuban-Americans elsewhere in the nation. There is much work to be done in preparing tabulations regarding income variables and conducting analyses on this important variable.
Further, while the focus of this paper has been on Cuban-Americans, the same techniques could be used to analyze Census Bureau data regarding other ethnic groups as well. Finally, I have left modeling of individuals’ outcomes based on demographic characteristics for future exploration, but it is clear that this data set is well suited to modeling income based on race, education and other characteristics. Clearly, the U.S. 2000 Population Census data provides the basis for the analysis of many potentially interesting relationships.
1. I would like to thank Jason Ost for assistance in preparing the statistical tabulations from the Census Bureau data and Lisandro Pérez for very helpful comments.
2. For detailed information on the questions in the short and long forms, see U.S. Census Bureau (1998).
3. Author’s calculation based on Census documentation.