Source of Data
This report includes data from the Census Bureau, the Bureau of
Labor Statistics (BLS), and the National Center for Health
Statistics (NCHS). The Census Bureau data in this report, which
cover a wide range of topics and years, were collected primarily
in the Current Population Survey (CPS), the Survey of Income and
Program Participation (SIPP), and the decennial census. The BLS
data are from the CPS. Data from the NCHS are from its vital
statistics registration system. The CPS deals mainly with labor
force data for the civilian noninstitutional population.
Survey estimates. The estimation procedures used
for CPS and SIPP data inflate weighted sample results to
independent estimates of the civilian noninstitutional population
of the United States by age, sex, race, and Hispanic/non-Hispanic
categories. These independent estimates are based on statistics
from decennial censuses; statistics on births, deaths,
immigration, and emigration; statistics on the size of the Armed
Forces; and, starting in 1994, an adjustment for undercoverage in
the census. The estimation procedure for 1994 used independent
estimates based on the 1990 decennial census; earlier data used
independent estimates based on the 1980 decennial census. (Data
in some sections are revised for earlier years. These revisions
are indicated in the relevant section.) This change in
independent estimates had relatively little impact on summary
measures, such as medians and percent distributions, but did have
a significant impact on levels. For example, use of the 1990
based population controls resulted in about a 1-percent increase
in the civilian noninstitutional population and in the number of
families and households. Thus, estimates of levels for 1994 will
differ from those for earlier years by more than what could be
attributed to actual changes in the population. These differences
could be disproportionately greater for certain population
subgroups than for the total population. The estimation
procedures used for CPS and SIPP data are discussed in more
detail in the publications cited at the end of the sections in
this report.
Reliability of Estimates
Since the CPS and SIPP estimates are based on samples, they may
differ somewhat from the figures from a complete census using the
same questionnaires, instructions, and enumerators. There are two
types of errors possible in an estimate based on a sample survey:
sampling and nonsampling. The standard errors provided in most Current Population Reports primarily indicate the
magnitude of the sampling errors. They also partially measure the
effect of some nonsampling errors in response and enumeration,
but do not measure any systematic biases in the data. Bias is the
difference, averaged over all possible samples, between the
estimate and the desired value. The accuracy of a survey result
depends on the net effect of sampling and nonsampling errors.
Particular care should be exercised in the interpretation of
figures based on a relatively small number of cases or on small
differences between estimates.
Sampling variability. Standard errors are
primarily measures of sampling variability, that is, of the
variations that occur by chance because of collecting a sample
rather than surveying the entire population. Standard errors are
not given in this report because of the wide range of topics
included and the wide variety of data sources. Standard errors
may be found in the publications that are noted at the end of
each section or by contacting the subject specialist.
Some statements in these publications may contain estimates
followed immediately by another number. For those statements, one
can add the number to the estimate and subtract the number from
the estimate to calculate upper and lower bounds of the 90-
percent confidence interval. For example, if a statement contains
the phrase "grew by 1.7 (+1.0) percent," the 90-percent
confidence interval for the estimate, 1.7 percent, would be from
0.7 to 2.7 percent.
Nonsampling variability. As in any survey work,
the results are subject to errors of response and nonreporting in
addition to sampling variability. Nonsampling errors can be
attributed to many sources, e.g., inability to obtain information
about all cases in the sample, definitional difficulties,
differences in the interpretation of questions, inability or
unwillingness on the part of the respondents to provide correct
information, inability to recall information, errors made in
collection such as in recording or coding data, errors made in
processing data, errors made in estimating values for missing
data, and failure to represent all units with the sample
(undercoverage).
Comparability with other data. Data obtained
from sample surveys and other sources are not entirely
comparable. This is due largely to differences in interviewer
training and experience and in differing survey procedures. This
is an additional component of error that is not reflected in the
standard errors. Therefore, caution should be used in comparing
results among these sources.
A number of changes were made in CPS data collection and
estimation procedures beginning in January 1994. The major change
was the use of a new questionnaire. The questionnaire was
redesigned to measure the official labor force concepts more
precisely, to expand the amount of data available, to implement
several definitional changes, and to adapt to a computer-assisted
interviewing environment. The March 1994 supplemental income
questions were also modified for adaptation to computer-assisted
interviewing, although there were no changes in definitions and
concepts. Due to these and other changes, one should use caution
when comparing estimates from data collected in 1994 with
estimates from earlier years. See the publications noted at the
end of each section for a description of these changes and the
effect they had on the data.
The April 1, 1990, census population was about 1.5 million less
than the estimate for the same date obtained by carrying forward
the 1980 census population with data on births, deaths, legal
international migration, and the net migration of U.S. citizens
across national boundaries. There are several possible
explanations for the difference, or "error of closure," including
a larger net underenumeration in the 1990 census, and
duplications and erroneous enumerations in the 1980 census. For a
detailed discussion, see J. Gregory Robinson, Bashir Ahmed,
Prithwis Das Gupta, and Karen A. Woodrow, "Estimating Coverage of
the 1990 U.S. Census: Demographic Analysis," 1991 Proceedings
of the Social Statistics Section, American Statistical
Association, pages 11-20.
This report includes data for five different population
universes: resident population plus Armed Forces overseas
(resident population as cited in official publications, does not
include Armed Forces overseas); resident population (census
universe); civilian population; civilian noninstitutional
population, plus Armed Forces living off post or with their
families on post (SIPP and March CPS universes); and civilian
noninstitutional population (CPS universe in months other than
March). The estimated size of the resident population including
Armed Forces overseas on July 1, 1994, was 260,651,000. The
estimated civilian noninstitutional population on July 1, 1994,
was 255,465,000 (table B-1). These population estimates are not
adjusted for estimated net underenumeration in the 1990 census.
However, for the first time, they incorporate a small increase
(8,418 persons) in the census-base population from count
resolution corrections processed through March, 1994.
While the civilian noninstitutional population has been adopted
as the universe for many sample surveys, the data in tables B-1
and B-2 are not consistent with results of current surveys
conducted by the Census Bureau through the end of 1993, including
the CPS, which were calibrated to 1980 census-based projections.
Current estimates for dates from January 1, 1994, onward, are not
consistent with the results of those surveys, including the CPS,
which are calibrated to projections that have been adjusted for
estimated net underenumeration based on the 1990 Post Enumeration
Survey.
The resident Armed Forces and the institutional population differ
greatly from the resident population in age-sex structure (table
B-2). On July 1, 1994, males 18 to 64 years old constituted 87.9
percent of the resident Armed Forces population, compared with
30.3 percent of the resident population, and females 65 years and
over constituted 36.6 percent of the institutional population,
compared with 7.6 percent of the resident population. However,
these two groups together (resident Armed Forces and
institutional population) accounted for only 2.0 percent of the
resident population. As a result, the civilian noninstitutional
population (which accounted for 98.1 percent of the resident
population) has an age-sex structure very similar to that of the
resident population. Similarly, the social and economic
characteristics of the resident Armed Forces and of the
institutional population could differ greatly from those of the
resident population despite relatively small differences between
the characteristics of the resident population and of the
civilian noninstitutional population.
Facebook
Twitter
Flickr
YouTube