In August 2001, the Cuban Government announced that it would conduct a census of population and housing from September 7 to September 16, 2002, the first census of the island’s population in over 20 years, and only the third such census since the revolutionary takeover of 1959.2 The census count was carried out and data were tabulated by the Oficina Nacional de Estadísticas (National Statistical Office, ONE). According to Cuba’s Minister of Economics and Planning José Luis Rodríguez, “there is no census of any country in the world with higher reliability [than Cuba’s 2002 census].”3
In November 2005, ONE released the long-awaited census results, giving them broad circulation by making them available in electronic from in its website.4 Researchers and journalists have seized upon these new data, reporting on such topics as basic demographic information about the Cuban population,5 the changing racial makeup of the island’s population, 6 its changing age structure,7 and internal migration patterns,8 among others.
In this paper, I compare selected demographic and social characteristics of Cubans in the island and Cuban- Americans in the United States in 2002. The comparison is based on Cuba’s 2002 census and statistics from the United States Census Bureau’s American Community Survey (ACS) for 2002, a data set that updates the U.S. decennial census. Naturally, some data in the two data sources do not lend themselves to direct comparison (e.g., wage information, or information about English proficiency of Cuban- Americans residing in the United States). Nevertheless, it has been possible to compare a limited number of indicators and, in so doing, to gain some interesting insights about similarities and differences in characteristics of Cubans in Cuba and Cubans in the U.S. diaspora at the turn of the century.
The paper begins with some background on the 2002 Cuban census and the ACS, and then compares data from each for selected demographic and social characteristics across different age groups. The paper explains how comparable data were developed, highlights points of interest, and makes a few suggestions for future research using Cuban and U.S. demographic and social data.
SOURCES OF DATA
Cubans in the Island: The 2002 Cuban Census9
As stated earlier, Cuba conducted a census of population and housing during the period September 7 through 16, 2002. The census was conducted entirely through personal interviews of heads of households or other adults. Interviewers were typically students enrolled in technical or professional high schools and at the university level, with instructors at these institutions serving as supervisors; in hard-toreach areas, local personnel was hired to conduct the census. Approximately 63,000 students and some 11,000 instructors were involved in data collection; in addition, over 20,000 persons were involved in data processing and other activities associated with the census. All personnel underwent significant training to prepare them for their activities associated with the census. Cuba published the results of the census 38 months after it was completed.
Cuban-Americans: The American Community Survey (ACS) 10
The American Community Survey (ACS) is an ongoing survey designed to collect detailed U.S. demographic and economic information previously collected through the decennial census. The ACS is a random sample, conducted by mail, with reminders sent out for those who do not immediately respond. The sample size is about 3 million households nationally, equating to about 1–in-40 addresses.
At the time of the decennial census, two types of forms are mailed out: “short” forms are sent to all housing units and “long” forms are sent to about 1– in-6 housing units. The “short” form serves to provide population counts, racial and ethnic characteristics, and age. The “long” form is the source of rich data on fertility, place of birth, English language proficiency, occupation, industry, household relationships, and income, as well as many other variables. These data take two to three years to tabulate and release. The ACS will eventually replace the “long” form in the 2010 decennial census. For data-users, this means that the rich data generated by the “long” form every decade will be updated annually at large geographies, and every 3 to 5 years at smaller geographies.
In this paper Cuban-Americans are defined as all respondents to the ACS who called themselves Cuban in response to the question about Hispanic ethnicity. Those respondents were further classified as either (1) U.S.-born (including Puerto Rico) Cuban-Americans or (2) foreign-born Cuban-Americans, on the basis of their response to the citizenship question.11 No distinction was made between naturalized citizens and non-citizens in the data.
In this paper I will refer to Cubans living in Cuba as “Cubans,” and self-identified Cubans living in the United States as “U.S.-born Cuban-Americans” or “foreign-born Cuban-Americans,” depending on their place of birth. This naming scheme is being introduced simply for the sake of clarity, and to avoid having to refer repeatedly to “Cubans in Cuba” or “Cubans on the island” and other over-long phrases.
The demographic and social variables compared in the next section were constructed from Cuban and U.S. official statistics that were developed not necessarily based on the same definitions. In order to construct roughly comparable indicators from the Cuban Census and the ACS, it was necessary to exercise some judgment.
• Child-Woman Ratio (CWR):12 The childwoman ratio is a rough measure of fertility where direct information is not available. The CWR is defined as:
where the numerator is the population of children under the age of 5, and the denominator is the number of women between 15 and 49 years old, inclusive. These women are considered to be the child-bearing age group. It was necessary to compute the CWR because the Cuban Census did not record fertility.
• Currently Married: The ACS and the Cuban census differ on how relationships are defined. The ACS differentiates between married, spouse present, and married, spouse absent. Here, they are combined to mirror the Cuban casado/a. Furthermore, the Cuban census includes the concept of unido/a consensualmente, defined as a person in a stable relationship between two people of opposite sex without any legal documentation. 13 While this approximates a common-law marriage in the United States, in the ACS, no such status is provided. Instead, couples “who live together (unmarried people, people in common-law marriages) were allowed to report the marital status they considered the most appropriate.”14 Because of these differences in definition, I chose to report only “Currently Married,” for all people 15 years of age and older. Those persons in the unido/a consensualmente respondents are classified with the non-married respondents. Additionally, in the born after 1959 tabulation, data were reported for those up to age 44, to conform to the given Cuban Census tabulation.
• Race: In the Cuban Census, enumerators were instructed to decide on the race of the respondent, and only to ask about other family members if they were not present and the enumerator had some doubt. Only three choices were given to enumerators, white, black, and mulatto/mestizo. In contrast, the ACS was conducted by mail, and all race categories were self-chosen. Not only that, but multi-racial responses were allowed, and in the most detailed version of the variable, 70 racial categories are listed, along with “other.” Since race is a concept that varies considerably across cultures, it seemed sensible to compare blanco with “white, alone,” negro with “black, alone,” and to avoid trying to create a mulatto/ mestizo variable out of the ACS data.
• Educational Attainment: Five general categories of attainment were constructed out of ACS data, corresponding loosely to primary school, middle school, high school, more than high school but less than a BA, and BA or more. The following corresponding categories were constructed out of the Cuban Census: primaria and lower, media básica (secundaria básica including the 10th year and obrero calificado), media superior (preuniversitario including the 13th year, técnico medio, and pedagogía nivel medio), up to three years of superior o universitaria, and four years or more of superior o universitaria. In the all ages tabulation, educational attainment was calculated for persons aged 25 to 59, inclusive. Although a range of 25 to 64 would have been more desirable, the presentation of the Cuban Census data did not allow this tabulation. In the born after 1959 tabulation, data were reported for those up to age 44, to conform to the given Cuban Census tabulation.
• Labor Force Status/Employment: The Cuban Census did not separate out civilians and members of the armed forces in its definition of employed persons, so members of the armed forces were included in the ACS employment tally to better compare across countries. The ACS employment status variable was compressed into three categories, employed, unemployed, and not in labor force. This will cause the employment totals and unemployment rates given to differ from the standard tallies produced for the civilian labor force. The unemployment rate for both Cuban Census data and ACS data was simply calculated as the number of unemployed persons divided by the sum of employed and unemployed persons. The labor force participation rate for both data sources was calculated as the sum of employed and unemployed persons divided by the sum of employed, unemployed, and those not in the labor force. In the all ages tabulation, labor force status and employment values were calculated for persons aged 16 to 64, inclusive.
Tables 1, 2, and 3 display selected demographic characteristics for the three major groups of interest: Cubans and Cuban-Americans, both U.S.-born and foreign- born. Table 1 presents characteristics for the three groups at all ages, Table 2 for the cohort born in 1959 and after, and Table 3 for the cohort that was 60 years or older in 2002. The cohort born after 1959 represents the children of the revolution, while the cohort 60 years of age or older represents those persons typically no longer in the labor force and drawing some sort of pension or social security payments.
Table 1 compares Cubans and Cuban-Americans, both U.S.-born and foreign-born, of all ages. Cubans in the island number 11,177,743, while Cuban- Americans number 1,355,898. This means that in 2002, Cuban-Americans made up around 11% of all Cubans in the island and in the U.S. diaspora.
The sex ratio does not differ greatly across the three groups, as all are almost evenly divided between males and females. The age structure, however, varies significantly. Foreign-born Cuban-Americans are clearly the oldest group, as 27.9% of this group is 65 or older, and fully 45% are 55 or older. Cubans are also much older than U.S.-born Cubans, as almost a fifth (19.7%) are 55 and older. The U.S.-born Cuban population is comparatively young, as fewer than 5% of U.S.-born Cubans are 55 and older, and over 60% are under 25. Meanwhile, a third of Cuba’s population is under 25. Close to half of Cubans and foreign-born Cuban-Americans are between 25 and 54, inclusive (46.6% and 48.1%, respectively), while closer to one third of U.S.-born Cuban-Americans fall into this age group (34.4%).
The child-woman ratio (CWR) also varies a great deal, probably in large part due to the age structure of the three groups. The older, foreign-born Cuban- American population has a very low CWR, while the younger U.S.-born Cuban-American population’s CWR is very high. The Cuban CWR is somewhere in between, as is the Cuban age structure. The share married for those 15 and older would also be affected by age, and this partly explains the lower share married for U.S.-born Cuban-Americans (39.5%) as well as the higher marriage rate for the foreign-born Cuban- Americans. The Cuban rate is just 35.1%, but it must be recalled here that the Cuban Census allows for an unido/a consensualmente status, which is a clue that many people live in stable common law-type marriages without legal standing.
Great differences can be observed in the racial makeup of Cuba and of the Cuban-American population. Both Cuban-American groups are over 80% white alone, while less than two-thirds of Cubans are white. Meanwhile, 10.1% of Cubans are black, but just a few percent of Cuban-Americans are black alone. The mulatto/mestizo category makes up onequarter of Cuba’s population, but only 13% of Cuban- Americans fall into all other races, and all mixed races. It should be noted, however, that Cuban mulatto/ mestizo totals cannot be compared directly with non-white, non-black totals for Cuban-Americans in the ACS.15 These international racial comparisons should be taken with a grain of salt for a number of reasons, among them: (1) the Cuban Census only provides three categories, while the ACS provides up to 71; and (2) because having enumerators assign race classifications differs from letting respondents choose their own. Not only that, but race categories are subjective, and differ across countries due to cultural differences. The low totals of “black, alone,” in the ACS, for instance, may be a response to the common usage of the term “black” in the United States, generally referring to non-Hispanic blacks. Still, the white and black totals at least tell us something about the racial composition of the Cuban and Cuban- American populations.
Of Cubans 25 to 59, inclusive, just under half have been educated up through media básica or less, about 9th grade (47.1%). Meanwhile, 97% of U.S.-born Cuban-Americans between those ages record educational attainment past the 9th grade. Foreign-born Cuban-American attainment falls between these two groups, as over one in ten have received a 9th-grade education or less (12.2%). Over one-third of U.S.- born Cuban-Americans have a BA or higher, while closer to one-fourth of foreign-born Cuban-Americans do, and closer to one-tenth of Cubans (38%, 22.9%, and 11.4%, respectively).
The Cuban labor force differs markedly from the Cuban-American one. The Cuban population age 16 to 64, inclusive, has a labor force participation that is 19 points lower than the Cuban-American group with the lowest labor force participation (57.4% for Cubans vs 76.5% for foreign-born Cuban-Americans). To check if this is partly a result of the Cuban retirement system, which provides for retirement for men at age 60 and for women at age 55,16 we can look at a tabulation from ACS data with a different age range. In table 1A, we use the same labor force status and employment variables but this time for women aged 16 to 54, inclusive, and men aged 16 to 59, inclusive, to reflect the retirement situation in the island. Restricting the ages of workers in this way raises the Cuban labor force participation rate almost 4 points (from 57.4% to 61.2%). Even with this change, the Cuban labor force participation rate is still significantly lower than the rate for all Cuban- American groups.
At 3.1%, the Cuban labor force has an unemployment rate than less than half of that of any of the Cuban- American groups examined. This very low Cuban unemployment rate is likely the result of some manipulation of statistics by the Cuban government. 17 U.S.-born Cuban-Americans are unemployed at a higher rate than foreign-born Cuban- Americans (8.9% vs. 7.3%).
Born in 1959 and Later
Table 2 displays characteristics of the cohort born in 1959 or after. This group was 43 years old and younger in 2002. As the rise of Fidel Castro to power dates to January 1, 1959, I think it is useful to consider the 1959 and younger population as a group representing the children of the revolution. Although the age structure was discussed above regarding Table 1, it bears mentioning that this group makes up over two-thirds of Cubans, over 90% of U.S.-born Cuban- Americans, and closer to one-third of foreign-born Cuban-Americans (36.1%). Half of all Cubans born after the Revolution are under 25, while fewer than one-fifth of foreign-born Cuban-Americans are (18.7%).
While the sex ratio for Cubans and for U.S.-born Cuban-Americans mirrors the all ages tally, foreignborn Cuban-Americans born after 1959 are just 44% female. This is likely a result of the lopsided character of the Mariel boatlift population. That migration wave was heavily male, and a man who was 20 in 1980—at the time of the boatlift—would be 42 at the time of the ACS, falling into the 1959 and younger category. The CWR is modified here, since ages are capped at 43. This has the effect of raising the CWR for all groups, since it removes women aged 44 to 49, who are less fertile than those aged 15 to 43. The same relationship seen in the all-ages population persists in this one, with foreign-born Cuban- Americans having the lowest CWR, U.S.-born Cuban-Americans having the highest, and Cubans having a CWR that is in between.
For those 15 and older, the share married is about three percentage points lower than the all-ages population for U.S. and foreign-born Cuban-Americans. The post-1959 Cubans, however, see a drop-off of around seven percentage points, suggesting that fewer of these Cubans are getting married, or that they are waiting until much later (after 43 years of age) to do it.
The racial composition of this group does not vary greatly from the all-ages group in Table 1.
Educational attainment for those 25 and over shows some differences from Table 1. In the post-1959 group, the share of Cubans who have completed primary school or less drops from 15.2% of Cubans of all ages to just 9%. The share that have completed from 10th grade up to finishing pre-universitaria jumps from 40.7% to 46.4%. The share with BA or better is essentially unchanged at 11.5%. U.S.-born Cuban-Americans in this age group show largely the same attainment as the all ages group. Foreign-born Cuban-Americans again have a level of attainment between that of Cubans and U.S.-born Cuban- Americans. The group in this age group has a slightly higher attainment than the all-ages group shown in Table 1. This is illustrated by the fact that just 1.2% of these Cuban-Americans have primary school education or less, compared to 4.2% of the all ages group.
Cubans in this cohort also have a lower labor force participation than Cuban-Americans. U.S.-born Cuban- Americans have a lower labor force participation than foreign-born Cuban-Americans, and a higher unemployment rate.
60 Years of Age and Older
As mentioned above, the retirement age for men in Cuba is 60 years of age and 55 for women, while workers in the United States can receive social security benefits as early as 62. It seemed logical to treat those 60 years and older as a separate group, already or close to being out of the labor force. It should be noted that this group accounts only for 3.3% of U.S.-born Cuban-Americans and 14.7% of Cubans.
Over one-third of foreign-born Cuban-Americans however, are 60 or older (36.9%). As expected, Cubans and foreign-born Cuban-Americans have a larger share of females than males, 52% and 53.6% respectively, reflecting the longer life-expectancy of women. U.S.-born Cuban-Americans however, are heavily male, as just 43.1% of this group is female.
When compared to the all-ages and post-1959 groups in Tables 1 and 2, this older group has a higher share married. Forty-four percent of Cubans, 53.8% of U.S.-born Cuban-Americans, and 57.7% of foreign-born Cuban-Americans are married.
This older group is whiter than the all ages and post- 1959 groups. In the case of Cubans, the increase in the share of whites (from 65.1% in the all ages group to 71.5% in the 60 and over group) comes almost entirely from the mulatto/mestizo group, as blacks make up around 10% of Cubans in Tables 1 and 3. In the case of foreign-born Cubans, there is little change in the share self-identifying as white.
Educational attainment in the 60 and older group shows an exaggerated version of the educational patterns seen in Tables 1 and 2. Over two-thirds of Cubans have a primary school education or less, while only a quarter of foreign-born Cuban-Americans list primary as their highest level of educational attainment (67.6% vs. 24.6%). Twenty-nine percent of foreign-born Cuban-Americans have post-high school education, while just 3.4% of Cubans can make the same claim.
This older cohort has a much lower labor force participation in both countries. Cubans have a labor force participation of just over 10%, while Cuban- Americans participate at a rate of just over 25%.
CONCLUSIONS AND AREAS FOR FURTHER RESEARCH
Cuban-Americans number 1.3 million, close to 11% of all Cubans in Cuba and the United States combined. The Cuban population is older than the U.S.- born Cuban-American population, and Cubans are far less likely to be married than U.S.-born and foreign- born Cuban-Americans. Perhaps as a result of those two factors, the Cuban population is less fertile than the U.S.-born Cuban-American population. The U.S.-born and foreign-born Cuban-American populations are whiter than the Cuban population, and the Cuban population is close to 10% black in all age groups examined. The Cuban population has the lowest educational attainment, and the U.S.-born Cuban-Americans have the highest educational attainment. Educational attainment is higher for those born after 1959 than for all ages and those 60 or older for all populations. Cubans have lower labor force participation than Cuban-Americans at all age levels, as well as lower unemployment.
The 2002 Cuban Census is a rich but flawed data source. As the first census of the Cuban population since 1981, it is of great importance for students of Cuban demography regardless of its quality. A judgment as to the quality of the data and of the datagathering process is beyond the scope of this paper. While the electronic release of data and documentation is a boon to researchers, greatly speeding analysis, the form of the data release is limiting, and influenced the variables I chose to analyze here. Data were released as 61 tables, which may seem to be a large number, but is actually quite sparse for a national census. In the area of educational attainment, for instance, only 3 tables were released, severely limiting the possibility for analysis. Even more limiting is the lack of microdata files in the Cuban Census, which would allow researchers to create their own cross-tabulations of the data.
The Cuban Census covers numerous topics not discussed in this paper. In particular, the release includes various tables on household dynamics and on housing, as well as data by territory and municipality for some fields. Other areas covered in the Cuban Census include workforce characteristics by occupation and industry. Researchers may want to work in these areas in the future. Some scholars have suggested research analyzing the Cuban and Cuban-American populations as complementary workforces in the context of a political and economic transition in the island that resulted in the free movement of labor between Cuba and the United States. This would entail, among other things, comparing the Cuban workforce’s labor experience and education with the Cuban-American workforce’s characteristics. Essential for all of this additional research would be release by Cuba of microdata from the 2002 census.
1. I’d like to thank Benigno Aguirre and Silvia Pedraza for thoughtful comments on an early version of this paper. All errors and omissions, however, are my own responsibility.
2. Revolutionary Cuba’s previous censuses were conducted in 1970 and 1982.
3. Iraida Calzadilla Rodríguez, “Cuba al detalle,” Granma (November 13, 2005). For concerns about the reliability of the census see, e.g., José Antonio Fornaris, “El censo más confiable del mundo,” www.cubanet.org (November 23, 2005); Michel Suárez, “El Censo de Oz,” www.cubaencuentro.com (November 29, 2005); Raúl Soroa, “Cuento de censoficción,” www.cubanet.org (December 5, 2005); and Oscar Espinosa Chepe, “Censo demorado y contradictorio,” parts I and II, www.cubanet.org (December 28 and 29, 2005).
4. Andrea Rodríguez, “Presentan los resultados del censo de población de Cuba,” Associated Press (November 12, 2005). Results of the census are at http://www.cubagob.cu/otras_info/censo/index.htm.
5. See, e.g., Pablo Alfonso, “Cuba en Cifras,” El Nuevo Herald (November 13, 2005).
6. E.g., “Cada vez más cubanos son mulatos,” www.cubaencuentro.com (November 14, 2005) and Oscar Espinosa Chepe, “Censo demorado y contradictorio (II y final),” www.cubanet.org (December 29, 2005).
7. E.g., “Cuba tendrá en 2020 la población más vieja de América Latina,” www.cubaencuentro.com (March 30, 2006).
8. “El 32% de los habitants de la capital cubana nació en otras provincias,” www.cubaencuentro.com (December 31, 2005).
9. This section is based on http://www.ccsr.ac.uk/cuba/cepde2004/censomultimedia/c_iii.htm.
10. This section is based on the Census Bureau website at www.census.gov, especially the press release “National Mailing of New American Community Survey Marks Historic Shift for Census Bureau” (January 10, 2005), available at http://www.census.gov/Press- Release/www/releases/archives/american_community_survey_acs/003349.html 11. Although it is the case that most foreign-born self-identified Cubans were born in Cuba, a person of Cuban descent born in Honduras, for example, who then immigrated to the United States could be both a foreign-born self-identified Cuban, and not born in Cuba itself. Still, it makes sense for this analysis of diaspora Cubans to include such a person in the group of Cuban-Americans. In any event, the number of pesons who fall into this category is rather small.
12. Gary Peters and Robert Larkin, Population Geography: Problems, Concepts, and Prospects, 4th Edition (Dubuque, Iowa: Kendall Hunt Publishers, 1993).
13. Taken from XII. Definiciones Básicas, Censo Cubano, 2002, available at http://www.ccsr.ac.uk/cuba/cepde2004/censomultimedia/ c_xii.htm CWR P F = − − 0 4 15 49 *1000
14. Taken from Marital Status definition, American Community Survey 2002 Subject Definitions, available at http://www.census. gov/acs/www/Products/PUMS/codelist2002.html
15. The Cuban census gives no explanation as to how someone whose ancestors were Chinese, but whose family had been in Cuba for generations would be categorized, for example. Presumably, enumerators would call such a person non-white, and non-black, but that would force them to be classified as mulatto or mestizo, which does not seem appropriate.
16. Lorenzo L. Pérez, “The Pension System of Cuba,” Cuba in Transition—Volume 8 (Washington: Association for the Study of the Cuban Economy, 1998), p. 522.
17. On this point see, e.g., Carmelo Mesa-Lago and Jorge Pérez-López, Cuba’s Aborted Reform (Gainesville: University Press of Florida, 2005), pp. 49–52.