Record Layout
ANNOUNCEMENT: INTERCENSAL USAGE NOTES/NEWS
Source: APDU Newsletter
Volume 17, No. 1
January/February 1993
Director Announces No Adjustment ofIntercensal Estimates
Barbara Bryant, Director of the Bureau of the Census, announced on
December 29 that the intercensal population estimates produced annually
by the Bureau will not be adjusted to correct for the estimated 1.6%
national net undercount in the 1990 census. This was one of the last acts of
Bryant, who left her position on January 20,1993. Harry Scarr, Deputy
Director, is now the senior official at the Census Bureau.
"This decision was difficult to make," Bryant said in her notice to the
Federal Renter. "It is the consensus of a committee of senior Census Bureau
statisticians and demographers that adjustment would improve the accuracy of
the 1990 census count at the national level, particularly for racial/ethnic
and owner/renter groups. Adjustment would also more accurately depict most
states' shares of the national population. The accuracy of the 1990 census
is important to the intercensal adjustment issue because the 1990 figure is
the startmg point for our yearly population estimates. However, because the
1990 total undercount is so small (less than 2 percent), it is impossible to
accurately adjust for the 44,055 substate areas for which population
estimates are produced."
Bryant noted that sponsors of federal surveys conducted by the Census
Bureau will be offered the option of calibrating their surveys to adjusted or
unadjusted population estimates beginnmg in 1993. These would be such surveys
as the Current Population Survey, the National Health Interview Survey, and
the National Crime Victimization Survey. These national surveys are
calibrated at large, aggregate levels where Census Bureau research shows
adjusted estimates are on average more accurate.
Bryant also pointed out that unadjusted estimates remain the official
population totals. Federal law requires that the Census Bureau issue official
totals for use in any federal program which allocates its funding based upon
population estimates. The Census Bureau provides official population
estimates to states and counties in every year between decenrialcensuses and
to subcounty areas every other year. Estimates of US population as of July 1,
1992, were released on January 1,1993.
* Reprinted with permission of the Association of Public Data Users (APDU)
For membership information contact:
Susan Anderson
(609) 258-6025 or
(609) 258-6052
ABSTRACT
Revised Estimates of the Population
of Counties by Age, Sex, and Race:
1980 to 1989
These data are revised estimates of the Population of the 3,141
counties in the United States as defined in 1990, by 5-year age
groups (age 0 to 4, 5 to 9,,,,,,,85 and over), sex (male,
female), and modified race (White, Black, other races) for each
year, July 1, 1980 through July 1, 1989. These data differ from
the Preliminary estimates of the Population of counties by age,
sex, and race available earlier this year. These revised
estimates are for counties as defined in 1990. These estimates
are also consistent with actual intercensal estimates for the
Nation by age, sex, and race and the actual intercensal estimates
for states by age and sex to be released later this year.
The revised estimates included in this release are developed by
interpolating between special census files for April 1, 1980 and
1990. The interpolations were adjusted to be consistent with:
1) revised intercensal estimates of the population of the nation
by age, sex, and race; 2) intercensal estimates of the
population of states by age-and sex; and 3) intercensal estimates
of the total population of counties.
The age and race data in the 1980 and 1990 Census data files
which were used as anchor points in this methodology were
modified to be consistent with other data sets. The 1980 census
data were further modified to be consistent with the 1990
definition of counties and the race modification Procedures
developed for 1990. The race Classifications were modified in
the 1980 and 1990 census files. The 1980 and 1990 census data
included persons classified as "Other race". These people were
not included in one of the specified racial categories listed on
the census form. The existence of the "Other race" category is
inconsistent with race categories defined by the Office of
Management and Budget in OMB Directive 15. In order to serve the
needs of the user community, it was necessary to assign each of
these persons to a specified race.
The age statistics for 1990 were modified to correspond with the
April 1, 1990 census date. A review of the detailed 1990 census
information indicated that respondents tended to provide their age
as of the date of completion of the questionnaire, not their age
as of April 1, 1990. In addition, there may have been a tendency
for respondents to round up their age if they were close to
having a birthday. The age data were modified to correspond to
the April 1, 1990 date.
The interpo1ations between the 1980 and 1990 modified census data
were adjusted to be consistent with: 1) the intercensal estimates
of the population of the nation by age, sex, and race; 2) the
intercensal estimates of the population of states by age and sex;
and 3) the intercensal estimates of the total population of
counties. The intercensal estimates for the population of the
nation by age, sex, and race, each state by age and sex, and
total population of each county are consistent with both the 1980
and 1990 censuses as enumerated. The latest set of l980-based
(postcensal) estimates totals and by age, sex, and race were
applied to the intercensal formula described in appendix I. All
intercensal State estimates are adjusted to be consistent with
the independent national intercensal estimates of population.
All intercensal county estimates are adjusted to be consistent
with the corresponding state estimates.
The data shown on this computer tape are unrounded. However, the
limitations of our methodology are such that we do not consider
these data to be accurate to the last digit.
METHODOLOGY
METHOD FOR PRODUCING INTERCENSAL ESTIMATES
OF THE POPULATION OF STATES: 1981-1989
This appendix describes the methodology used to derive the
intercensal estimates of the resident population of States,
divisions, and regions for July 1, 1981 to 1989.
The intercensals are consistent with the 1980 Census counts as
published in the U.S. Census Summary, PC80-l-B1 and the 1990
census modified counts. The latest set of 1981-90 postcensal
(1980-based) State estimates were used in the intercensal formula
described below. The sum of the intercensal state estimates for
each year was adjusted to be consistent with independent national
estimates of tne intercensal population estimates.
Algebraically, the intercensal procedure can be described as:
t/10
P = Q (P / Q )
t t 10 10
The formula above is describeed in terms of the following
notations:
= time (years) elasped since
April 1, 1980, 0 < t < 10
- -
Q = postcensal estimaate at time t
t
P = intercensal estimate at time t
t
so that
Q = P = actual census count as
o o of April 1, 1980
P = actual census count
10 as of April 1, 1990
Q = postcensal estimate for
10 April 1, 1990
This procedure was also used in developing the annual intercensal
population estimates for the United States.
RECORD LAYOUT
General Information
This file contains revised estimates of the population of the
3,141 counties (1990 geography) in the United States by five-year
age groups (ages 0-4, 5-9, .85 and over), sex (male, female) and
modified race (white, black, other race) for each July I date
1980 through 1989.
Record Layout
Contents Columns
Year (2 digits, 80.....89) 1-2
TIPS (State/County) 3-7
Race/Sex code (1 white male 8
2 white female
3 black male
4 black female
5 other male
6 other female)
Blank 9
Age 0=4 10-17
5-9 18-25
10-14 26-33
15-19 34-41
20-24 42-49
25-29 50-57
30-34 58-65
35-39 66-73
40-44 74-81
45-49 82-89
50-54 90-97
55-59 98-105
60-64 106-113
65-69 114-121
70-74 122-129
75-79 130-137
80-84 138-145
85+ 146-153
