U.S. Population Estimates by Age, Sex, Race, Hispanic Origin, and Nativity: 1990 to 1999 Source: Population Estimates Program, Population Division, U.S. Census Bureau, Washington, D.C. 20233 Internet Release date: April 11, 2000 TABLE OF CONTENTS I. DATA FILE DESCRIPTIONS II. DATA FILE RECORDS A. Record Layout B. Explanation of Data Fields III. POPULATION UNIVERSE AND TIME SERIES IV. THE ESTIMATES BASE POPULATION V. POSTCENSAL ESTIMATION OF RESIDENT POPULATION A. Estimation of Population By Sex, Race, and Hispanic Origin B. Estimation of Components of Population Change C. Estimation of Population By Age D. Estimation of Population By Quarter I. DATA FILE DESCRIPTIONS The National Population Estimates by Nativity page on the U.S. Census Bureau web site links to two files of detailed national population estimates data. One file contains quarterly postcensal population estimate data between April 1, 1990 and July 1, 1998 for native United States residents. The other file contains quarterly postcensal population estimate data over the same time period for foreign-born United States residents. Each series contains estimate files that cover the nation (50 states plus the District of Columbia) by single year of age, sex, race, and Hispanic origin. The quarterly postcensal national estimates by nativity are available only on the World Wide Web. There are no corresponding paper or electronic product. Neither series of estimates are adjusted for net undercoverage found in either the 1990 Decennial Censuses, but they do contain adjustments for census count question resolution corrections and net undercoverage found in the test sites for the 1995 test censuses and 1998 census dress rehearsals. All estimates are consistent with currently published annual estimates of the population prior to 1980 by age, sex, and race, as they appear in Current Population Reports, Series P-25, Numbers 311, 519, and 917. These national nativity estimates are preliminary and subject to future revision. Descriptions of the national nativity estimates files follow below. Note that the files are stored in ASCII format. File Name File description FOR9099Q.TXT Quarterly postcensal foreign-born resident population estimates, April 1, 1990 to July 1, 1998, by age, sex, race, and Hispanic origin (34 reference dates, 3468 records, 811,512 bytes) NAT9099Q.TXT Quarterly intercensal native resident population estimates, April 1, 1990 to July 1, 1998, by age, sex, race, and Hispanic origin (34 reference dates, 3468 records, 811,512 bytes) II. DATA FILE RECORDS A. Record Layout The 1990 quarterly foreign-born and native resident population estimate files have an identical layout. Each record has a total length of 222 characters. All data fields are right-justified. Location Type Data 1-2 Character Series 3-4 Numeric Month 5-8 Numeric Year (last 2 digits) 9-11 Numeric Age (years) 12 (blank) (blank) 13-22 Numeric Total population 23-32 Numeric Total male population 33-42 Numeric Total female population 43-52 Numeric White male population 53-62 Numeric White female population 63-72 Numeric Black male population 73-82 Numeric Black female population 83-92 Numeric American Indian, Eskimo, and Aleut male population 93-102 Numeric American Indian, Eskimo, and Aleut female population 103-112 Numeric Asian and Pacific Islander male population 113-122 Numeric Asian and Pacific Islander female population 123-132 Numeric Hispanic male population 133-142 Numeric Hispanic female population 143-152 Numeric White, non-Hispanic male population 153-162 Numeric White, non-Hispanic female population 163-172 Numeric Black, non-Hispanic male population 173-182 Numeric Black, non-Hispanic female population 183-192 Numeric American Indian, Eskimo, and Aleut, non-Hispanic male population 193-202 Numeric American Indian, Eskimo, and Aleut, non-Hispanic female population 203-212 Numeric Asian and Pacific Islander, non- Hispanic male population 213-222 Numeric Asian and Pacific Islander, non- Hispanic female population Within each file, the records are first sorted by the reference date (Month-Year) in chronological order. For each reference date, the first record lists the population counts for all ages combined. The remaining records list the population counts by single year of age in ascending order. B. Explanation of Data Fields Series: Each national population estimate record has the series designation "9P", which represents the year of production (1998) and the fact that they are postcensal estimates. Month: The reference dates for quarterly estimates coincide with the first day of the first month of the quarter. For the foreign-born and native national population estimates, the reference dates have month codes of "1", "4", "7", and "10", which represent the first day of January, April, July, and October, respectively. Year: Each year is identified by four digits. For example, a code of "1999" indicates the year 1999. In previous releases, year has been identified by the last two digits. Age: The first record for each reference date has an age code of "999", which lists the population counts for all ages combined. Each age code that follows represents a single year of age, except for "100", which lists the population counts for people 100 years of age and over (centenarians). Sex, Race and Hispanic Origin: Each record lists population counts for 21 different sex-race-Hispanic origin categories. The first category shows the total population for the listed age. The remaining 20 categories list the male and female populations, respectively, for the following racial and/or Hispanic origin categories: Total (All races) White Black American Indian, Eskimo, and Aleut Asian and Pacific Islander Hispanic origin (of any race) Non-Hispanic White Non-Hispanic Black Non-Hispanic American Indian, Eskimo, and Aleut Non-Hispanic Asian and Pacific Islander III. POPULATION UNIVERSE AND TIME SERIES Quarterly estimates for the period April 1, 1990 and forward are postcensal estimates, based primarily on the 1990 Decennial Census enumeration and estimates of the population change from the census date to the reference dates of the estimates. Because the 1990 Decennial Census is the only comprehensive enumeration of the United States population in the series, there is no error of closure. Estimates of the United States resident population include people who reside in the 50 States and the District of Columbia. They exclude residents of the Commonwealth of Puerto Rico, and residents of the outlying areas under United States sovereignty or jurisdiction (principally American Samoa, Guam, Virgin Islands of the United States, and the Commonwealth of the Northern Mariana Islands). The definition of residence conforms to the criterion used in the 1990 census, which defines a resident of a specified area as a person "usually resident" in that area. Estimates of resident population exclude the United States Armed Forces overseas, as well as civilian United States citizens whose usual place of residence is outside the United States. IV. THE ESTIMATES BASE POPULATION While the 1990 Decennial Census defines most of the base population for these estimates, the total resident population for April 1, 1990 (248,790,925) does not agree with the 1990 census figure as originally published (248,709,873). The difference, as reflected throughout the entire postcensal estimates series, arises from the sources described below: (1) The estimates include census count question resolution corrections determined through the Census Quality Review process. These corrections of census tabulations, which are not related to underenumeration, came after the publishing of the initial 1990 census results. (2) The estimates include population adjustments associated with the underenumeration that resulted in the eight test sites from the 1995 Census Test. These test sites include Oakland, California; Paterson, New Jersey; and six Louisiana parishes (Bienville, De Soto, Jackson, Natchitoches, Red River, and Winn). They also include population adjustments associated with the underenumeration in the 1998 census dress rehearsals conducted in Sacramento, California and Menominee County, Wisconsin. An administrative requirement for local estimates to conform to these adjusted counts necessitated a revision of the national population definition to include the adjustments. With the exception of the 1995 test census sites, and 1998 dress rehearsal sites, the population does not include adjustments for underenumeration. In particular, the sample-based "dual system" estimates of underenumeration derived from the 1990 Decennial Census Post Enumeration Survey, included in the results of many federal surveys, are not included in these estimates. Aside from the changes affecting total population, modifications were made in the distribution of the population by age and race. These modifications brought the definition of age and race into conformity with definitions used for data from other sources, such as vital statistics. The age distribution of the estimates base population, while census-based, differs with the distribution shown in the 1990 census publications. Such modifications brought the definition of age into conformity with definitions used for data from other sources, such as vital statistics. The aim was to correct biases in census age tabulations that resulted from displacement of age reporting from the reference date of the census. In 1990 census publications, age is based on respondents' direct reports of age at last birthday, with some editing for age misstatement. This definition proved inadequate for postcensal estimates, as it was apparent that many respondents reported their age (even if reported correctly) at the time of completion of the census form or interview by an enumerator. These reported ages usually occurred up to several months after the April 1 reference date, resulting in an upward biased age. Modification was based on a re-specification of age, for most individual respondents, according to their year of birth. Age was derived from year of birth by allocating date of birth to the first quarter and last three quarters of each year, subtracting year of birth from 1990 for those born before April 1, and from 1989 for those born after April 1. The allocation was based on a historical series of registered births by month. As with age, the census-based race distribution of the estimates base population differs with the distribution shown in the 1990 census publications. Such modifications brought the definition of race into theoretical conformity with other sources, such as vital statistics. The modification conformed to the definition of race specified in Office of Management and Budget Directive 15. In the 1990 census, a substantial number of people (roughly 9.8 million) did not specify a racial group that could be classified as any of the White, Black, American Indian, Eskimo, Aleut, Asian, or Pacific Islander categories on the census form. A large majority of these respondents were of Hispanic origin, based on their response to a separate Hispanic origin question on the form. Most wrote in "Hispanic" or their Hispanic origin type (such as "Mexican" or "Puerto Rican") as their race. People of unspecified race were allocated to one of the four tabulated racial groups (White; Black; American Indian, Eskimo, and Aleut; Asian and Pacific Islander) based on their response to the Hispanic origin question. While this modification does not ensure that race reporting will be consistent among data sources, it enables the use of varying sources of data in the estimation process, with appropriate adjustment of categories. Nativity, unlike any other variable estimated here, is not reported on the "short form" census questionnaire, but was reported on the "long form" census questionnaire received by roughly one person out of every six. Nativity is defined by the question on U.S. citizenship, not on actual place of birth. A person is "native" if he or she was a U.S. citizen at the time of birth. Natives may thus be either born in the United States, or born abroad with at least one U.S. citizen parent. Foreign-born persons are those who are not native, including naturalized citizens and non-citizens residing in the United States. Because the distribution of the population by age, sex, race, Hispanic origin, and nativity was, of necessity, based on a sample, the estimated totals by age, sex, race, and Hispanic origin did not match precisely the corresponding cross-categories in the base population by age, sex, race, and Hispanic origin. This situation was corrected by a proportional adjustment of each age, sex, race, Hispanic origin, and nativity category so that natives and foreign-born summed to the base population category defined by the other variables. V. POSTCENSAL ESTIMATION OF RESIDENT POPULATION A. Estimation of Population By Sex, Race, and Hispanic Origin Estimates of the United States population by 32 combinations of sex (male, female), race (White; Black; American Indian, Eskimo, and Aleut; Asian and Pacific Islander), Hispanic origin (Hispanic, Non-Hispanic), and nativity (native, foreign-born) were derived quarterly by updating the resident population enumerated in the 1990 census through the components of population change. The following formula was applied to update each group: (1) 1990 enumeration of resident population, (2) + births to U.S. resident women, (3) - deaths to U.S. residents, (4) + net international migration, (5) + net movement of U.S. Armed Forces and civilian citizens to the United States. It should be noted, however, that births are consistently zero for the foreign-born population, since people born within the United States becomes a native. B. Estimation of Components of Population Change While the estimates base population was described in the previous section as being primarily a consequence of the 1990 Decennial Census, the components of change - births, deaths, and both major classes of net migration - are estimated from administrative sources. (1) Registered births to United States resident women are estimated from tabulations of data on birth certificates, as supplied by the National Center for Health Statistics. The primary source for these birth data can be found in the National Center for Health Statistics Monthly Vital Statistics Report, Volume 47, Number 18, "Births: Final Data for 1997," and similar reports for other years. Generally, it is assumed that the coding of race and Hispanic origin of mother on birth certificates is consistent with the reporting of race and Hispanic origin of children in the census. Because of apparent serious discrepancies in the reporting of race for the Hispanic origin population between the census and birth certificates, however, cross-categories of race and Hispanic origin are built on the assumption that rates of childbearing among Hispanic women do not vary by race. (2) Registered deaths to United States residents by sex and race are estimated from tabulations of data on death certificates, as supplied by the National Center for Health Statistics. The primary source for this death data can be found in the National Center for Health Statistics National Vital Statistics Report, Volume 47, Number 19, "Deaths: Final Data for 1997," and similar reports for other years. It is assumed that reporting of race on death certificates is consistent with reporting of race in the census. Because of the number of decedents of unknown Hispanic origin is large, age-specific Hispanic death rates are calculated for a large proportion of the states where reporting of Hispanic origin is complete. These rates are then used to estimate Hispanic deaths for the entire U.S. population. For each cross-category of race and Hispanic origin, it is assumed that native and foreign-born age-specific death rates are the same. (3) The international net migration components are based on a variety of administrative sources and analytic estimates. The Immigration and Naturalization Service supplies data on legal immigrants. The Office of Refugee Resettlement supplies data on people admitted to the United States as refugees. Net migration from Puerto Rico to the United States is projected at a constant level, based on an estimate of annual net migration out of the Commonwealth during the 1980s, as developed by the Census Bureau's International Programs Center. Subtracting legal immigrants to the island during the 1980s, and assuming that undocumented migration and the emigration of legal residents from the island net to zero, yields an estimate of net migration from Puerto Rico to the United States. The (presumably small) balance of migration between the United States and other outlying areas under United States jurisdiction is assumed to be zero. Because international migration data sources generally code country of birth or citizenship rather than race and Hispanic origin, tabulations of country of birth by race by Hispanic origin from the 1990 census are used to impute race and Hispanic origin of migrants. A proportion of the net migration from Puerto Rico to the United States, as based on the 1990 census, is assumed to be native to the United States (including persons born in Puerto Rico). Two numerically substantial types of international migration for which reliably accurate and current data are unavailable are undocumented immigration (both in and out of the United States) and the permanent emigration of legal residents. Undocumented immigration refers to the net increase of the undocumented population due to change in residence across the national border by aliens, either by unauthorized entry or by overstaying a temporary (non-immigrant) visa, discounting those that subsequently immigrate legally. For undocumented immigration, we make an allowance of 225,000 net migration per year, which is constant throughout the postcensal estimating period. The race and Hispanic origin of this flow is based on distributions by country of birth estimated by the Immigration and Naturalization Service, and imputed by tabulations of race by Hispanic origin by country of birth from the 1990 census. All undocumented migration is assumed to be foreign-born. Calculating the emigration of legal residents requires calculating native and foreign-born emigration separately. Foreign-born emigration counts incorporated results of research carried out at the Census Bureau and published as Population Division Technical Working Paper #9, "Estimates of the Emigration of the Foreign-Born Population: 1980 to 1990," authored by Bashir Ahmed and J. Gregory Robinson. These estimates for the 1980s were used to calculate rates of emigration for the foreign-born population, which were projected through the 1990s. Legal native resident emigration counts relied on an allowance of 48,000 per year, based on Population Division Technical Working Paper #10, "Estimation of the Annual Emigration of U.S. Born Persons by Using Foreign Censuses and Selected Administrative Data: Circa 1980," authored by Edward Fernandez. Also included was an allowance for the net migration of non-refugee temporary residents (most of which are foreign students and scholars) who would be enumerated in the decennial census as residents of the United States. This allowance is intended to preserve the age, sex, race, and Hispanic origin distribution of the population as estimated for the 1990 census date. (4) The temporary movement of federally affiliated United States citizens consists mostly of movement of the active duty Armed Forces and Armed Forces dependents between the United States and overseas. Also included is the movement of civilian federal employees and their dependents. We estimate the migration of federally affiliated citizens by observing trends in the number of personnel abroad, from three sources: a) Armed Forces overseas strength statistics are supplied by the five branches of the Armed Forces in the Departments of Defense (Army, Navy, Marines, and Air Force) and Transportation (Coast Guard). b) The number of military dependents residing overseas is published by the Department of Defense, for four branches of the Armed Forces (excluding Coast Guard). These data come from the Department of Defense, Directorate for Information, Operations, and Reports, "Worldwide Manpower Distribution by Geographic Area." c) The number of civilian Federal employees overseas is supplied by the Office of Personnel Management. This number is inflated to include dependents. These data come from the United States Office of Personnel Management, Federal Civilian Workforce Statistics, "Employment and Trends" series. The inference of net migration from this population to the United States follows a reverse component-of-change logic applied to the federally affiliated overseas population. The net increase in the overseas Federal population, minus net recruits to this population from abroad (for example, residents of outlying areas joining the Armed Forces), minus births, plus deaths, provides an estimate of the net migration to the overseas Federal population from the United States. The negative of this estimate is the estimated migration of federally affiliated citizens to the United States from overseas. Distributions of the Armed Forces overseas by race, Hispanic origin, and nativity are supplied by the Department of Defense, Defense Manpower Data Center. These distributions are imputed to the balance of the overseas federally affiliated population, most of which are Armed Forces dependents. The reverse component-of-change logic is applied to each race-origin cross-category. The net migration of United States citizens not affiliated with the Federal government, as well as the temporary movement of students, scholars, and embassy personnel, are tacitly assumed to be zero, although these movements do, in fact, represent a change of residence in and out of the United States according to the residence definition used in the estimates. C. Estimation of Population By Age (1) The inflation-deflation method. Once the postcensal population estimates by sex, race, and Hispanic origin are computed, the age distribution of each sex-race-Hispanic category is estimated by the inflation-deflation method. This method seeks to allow each age group in the census base population to advance one year of age for each year after the census, reflecting the effects of births, deaths and migration, while preserving - without aging - the age pattern of census undercoverage and age misreporting estimated for the base population. The method is described in detail in Current Population Reports, Series P25, Number 1095, "U.S. Population Estimates, by Age, Sex, Race, and Hispanic Origin: 1980 to 1991," pp. xvii-xviii. (2) Components of change by age. Deaths to United States residents by age (single years through age 69, and the total of ages 70 and over), sex, and race are based on official tabulations of deaths by single year of age from the National Center for Health Statistics. The age distribution of deaths to residents 70 years of age and over are based on rates computed from life tables, and proportionally adjusted to be consistent with the National Center of Health Statistics' total deaths to people 70 and over. For the Hispanic population by sex and race, the 1990, 1992, and 1994 Hispanic life tables are applied to produce Hispanic deaths by age, sex, and race. Non-Hispanic deaths are then derived as total minus Hispanic, within categories of age, sex, and race. Deaths by nativity within age, sex, race, and Hispanic origin categories are based on the assumption of equal rates of mortality between native and foreign-born persons in the category. For legal alien immigrants admitted by the Immigration and Naturalization Service, age and sex is available on the individual immigrant records. For refugee categories, the age distribution is based on distributions of refugees of the same country of birth adjusting to permanent immigrant status through the Immigration and Naturalization Service, also from the individual immigrant records. We estimated the age distribution of federally affiliated United States citizens by the method of cohort survival, using age distributions of Armed Forces overseas supplied by the Department of Defense's Defense Manpower Data Center, as well as distributions of overseas Armed Forces dependents from the 1970 census, and overseas civilian federal employees from the Office of Personnel Management. D. Estimation of Population By Quarter All of the postcensal population detail shown in these files was produced for monthly reference dates, although the calendar quarter was the basic time interval for estimating distributions of components of change by age, sex, race, and Hispanic origin. For the most part, the data sources for components of change provide monthly data, to the extent that the data are final. For some preliminary data, the monthly series incorporated projections of seasonal distributions from final data for earlier years. Data from the Immigration and Naturalization Service and Office of Refugee Resettlement on legal alien immigration and refugees, respectively, were available by month. Most Armed Forces strength data by branch of service, used to infer movement of Armed Forces from overseas, were available by month, whereas data on Armed Forces dependents, and all Armed Forces demographic detail before 1993, were available by quarter. No seasonal information was available for undocumented immigration, emigration of legal residents, net migration from Puerto Rico to the United States, and net migration of non-refugee temporary residents, so no seasonality was assumed. For births, deaths, and legal alien immigrants, quarterly distributions by age, sex, race, and Hispanic origin were independently derived. For other components, sex, race, and Hispanic origin distributions were imputed to quarterly data; either from the current calendar year or fiscal year, or on the basis of a projection from a past year. The procedure for estimating the population by quarter, once the quarterly components of change were estimated, followed the procedure described above. The single-year age distribution of the base population was further distributed to quarter-year of age in proportion to registered births in the birth year of each cohort. The inflation-deflation method was employed to the quarter-year of age population, as updated through components of change by age, sex, race, and Hispanic origin.