gerencia y administración censal. ignacio velez...
TRANSCRIPT
Gerencia y Administración Censal.
Ignacio VELEZ
Departamento Administrativo Nacional
Estadística
COLOMBIA
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
INTRODUCCION
Este documento ilustra el manejo del operativo censal del XV
Censo Nacional de Población y IV de Vivienda de Colombia,rea
lizado a partir del 15 de octubre de 1985, en particular, el
manejo administrativo y logístico del proyecto. La trasmi-
sión de esta experiencia es una contribución para mejorar en
el futuro la ejecución de los censos en América Latina.
ANTECEDENTES
2.1 Historia Censal.
Desde 1770 se han realizado 19 censos oficiales. El ú_l
timo se llevó a cabo en 1973 y fue publicado en 1981. -
Este censo no fue aprobado.
2.2 Usos legales de los censos.
La población registrada en el último censo aprobado es
la base para calcular el número de congresistas, dipu-
tados y concejales que cada jurisdicción político-admi-
nistrativa debe tener. Además, sirve para fijar la
parte del Presupuesto Nacional que se asigna a las re-
giones .
2.3 División Político-Administrativa de Colombia.
Colombia está dividida en un distrito especial, 23 de-
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
64
partamentos, 4 intendencias y 5 comisarías. En estas
últimas, que ocupan casi la mitad del territorio, se
encuentra cerca del 2% de la población. Las grandes
divisiones territoriales constan de 990 municipios,
1,818 corregimientos, 3,919 inspecciones de policía,
1,797 caseríos y grandes extensiones con población dis-
persa .
2.4 Funciones del Departamento Administrativo Nacional de
Estadística (DANE).
El DANE elabora las investigaciones estadísticas necesa
rias para la formulación de planes de desarrollo y en
particular realiza los censos de población y vivienda.
Tiene tres direcciones generales: la Dirección General
de Información Técnica, la Dirección General de Análi-
sis Socioeconómico y la Dirección General de Procesa-
miento de Datos.
Para el trabajo del DANE, el país se divide en seis regio
nes: Central, Norte, Noroccidental, Nororiental,Centrooc
cidental y Suroccidental. Las oficinas regionales tie
nen Jefes de Zona que se encargan de recolectar informa
ción en los municipios de cada departamento.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
65
3. XV CENSO NACIONAL DE POBLACION Y IV DE VIVIENDA: CENSO 85.
3.1 Orígenes del CENSO 85.
En 1979, el DANE propuso la realización de un nuevo cen
so de población y se consideró crear una Dirección Ge-
neral dedicada exclusivamente a ello. Asimismo, se
consideró complementarlo con una muestra. Esta pro-_
puesta fue rechazada, entre otras razones porque el Cen
so de 1973 no se había publicado aún. En 1983 se reto
mó la idea y fue aprobada el 24 de junio del mismo año.
En diciembre se crearon la Junta Nacional del Censo y
el Comité Técnico. En 1984 se creó la Dirección Gene-
ral del Censo.
3.2 Objetivos.
Para que una organización funcione bien debe tener obje
tivos claros y bien definidos, para el CENSO 85 se esta
blecieron en forma muy precisa los siguientes: saber
cuántos somos y dónde vivimos, para construir un marco
muestral. Además se estipuló que el éxito del censo
dependía de la oportunidad, de la alta cobertura y de
la conflabilidad.
Un censo oportuno significaba entregar datos preliminares
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
66
el 20 de diciembre de 1985 y resultados definitivos ?
irás tardar el 31 de julio de 1986 . La oportunidad de
los datos censales era crucial debido a la experiencia
de 1973. Se fijó como meta una cobertura del 98%. Es-
to era muy importante debido a que en 1973 se logró
una cobertura promedio de 90% , con sitios de 42% y de
97.2%. Un censo debe producir información confiable y
precisa. Esto significaba que no se podía distorsio-
nar la información en ninguno de los procesos.
Para lograr estos objetivos se adoptaron estrategias
que garantizaran el éxito del proyecto. Algunas de
ellas son:
a) utilizar la planeación estratégica y detallada;
b) utilizar la tecnología más avanzada;
c) realizar censo sin inmovilización; y
d) utilizar las técnicas de muestreo.
Con la planeación estratégica se definieron los ob jeti_
vos generales del proyecto, las políticas, metodologías
y estrategias que permitieran precisar los procesos si
guientes para el desarrollo del CENSO 85. La planea-
ción detallada desagregó el proyecto en otros de menor
envergadura y estos, a su vez, en subproyectos con los
siguientes criterios:
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
67
1. Que fuera posible la programación y el control en
forma eficaz y eficiente.
2. Que cada subproyecto tuviera un responsable direc
to.
EL CENSO 85 se descompuso en tres eventos: precensales,
censales y postcensales.
3.3 Eventos precensales.
Estos se agrupan en grandes categorías:
a) organización censal;
b) cartografía y recuentos de vivienda;
c) formularios, pruebas piloto y censos experimentales;
d) preparación de infraestructura para la recolección;y
e) publicidad.
3.3.1. Organización censal.
La organización censal consta de la organización censal
central y de la organización censal para la recolección,
3.3.1.1, Organización censal central.
Todo el personal del CENSO 85 estuvo dedicado en forma
exclusiva a la realización del proyecto. De esta mane-
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
68
ra, el grupo tuvo responsabilidades específicas, con -
metas muy claras y precisas que cumplir en fechas est j_
pulada 5.
Tener una organización separada del DANE para Ki pla-
neación y ejecución del CENSO 85, garantizó que las
actividades normales del DANE no sufreran menoscabo.
La autonomía relativa del grupo, no sólo en cuanto a
su separacrón organizaclona 1 y física, sino a la índe-
pendencia en el desarrollo del trabajo, fue factor de-
cisivo para el logro de los objetivos. Al configurar
un grupo "extraño" al DANE se perdía parte de la expe-
riencia en términos institucionales. Para reducir este
efecto, se realizaron "empalmes", en particular en las
áreas de cartografía, muestras, finanzas, evaluación y
procesamiento de datos, a través de las respectivas uní
dades del CENSO 85 y las direcciones generales del DANE
En marzo de 1984 se crearon las unidades encargadas del
CENSO 85 así: Operativa para diseñar, ejecutar y con-
trolar la recolección; Planeación y Control para di se-
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
69
ñar los mecanismos de plane
rentes eventos censales; Se
proporcionar soporte logíst
municación y Educación para
cargado de la recolección y
Recursos Humanos para defin
mas de reclutamiento del pe
tos para diseñar la estruct
y procesamiento de datos, y
y las cantidades de equipo
manejar los recursos f inane
para allegar recursos adici
a través de entidades
Este grupo no había t
sos, por lo tanto su
censos de Colombia y
ación y control de los dife
rvicios y Suministros para
ico a todo el proyecto; Co~
capacitar al personal en-
promoclonar el proyecto;
ir los perfiles y los siste
rsonal; Procesamiento de Da
ura del sistema de captura
ir las características
rio; Financiera para
y Recursos Especiales
a los ya mencionados,
e internacionales,
a experiencia en cen-
a fue estudiar los
ses .
nacionales
enido ningún
primera tare
de otros paí
def in
necesa
ieros;
onales
Entre agosto de 19 84 y junio de 1985 la estructura or-
ganizacional sufrió cambios que produjeron una organi-
zación con divisiones funcionales. La forma como se
agrupó a las unidades respondió al grado en que sus fun
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
70
clones se complementaban, al tipo de proyectos que te-
nían a cargo y al liderazgo natural surgido en el gru-
po. Estos cambios respondieron a las necesidades de
cada una de las etapas vividas en el CENSO 85. En otras
palabras, se contaba con una estructura organizacional
dinámica y flexible. Para garantizar la independencia
de la muestra de cobertura, este operativo y la Unidad
de Muestras, pasó a depender de la Dirección General -
de Análisis Socioeconómicos (DANAL). En enero de 1986
se constituyó formalmente la División de Metodología y
Evaluación.
En marzo de 1984 se contrataron 12 personas; el número
máximo fue de 172 y se alcanzó en octubre de 1985. El
60% tenía título profesional; de éstos, el 8.7% tenía
posgrado. El 40% restante era personal auxiliar (tec-
nólogos y estudiantes universitarios) y soporte admi-
nistrativo. El 42% del personal era del sexo femenino.
La gran mayoría provenía del sector privado.
Durante todo el tiempo se contó con asesoría internado
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
71
nal y local sobre organización censal, procesamiento
de datos, cartografía y control de calidad. Este últi_
mo, relacionado no sólo con la grabación de datos, si-
no con los procesos logísticos para garantizar la ade-
cuada y oportuna recolección de los datos.
El Jefe del DANE y el Director del CENSO 85 trabajaron
perfectamente coordinados e identificados en políticas
y objetivos. Aunque aquel tuvo mucha ingerencia e in-
terés en el proyecto, sus intervenciones fueron básica
mente orientaciones de tipo general y de f i j ación de
políticas. Esto significó que el grupo contó con li-
bertad y apoyo incondicional para desarrollar el traba
jo. Asimismo, las unidades tuvieron suficiente autono
mía para el desarrollo del trabajo. Sin embargo, fue-
ron inflexibles en mantener las políticas y decisiones
que se consideraban necesarias para garantizar el éxi-
to del CENSO 85.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
72
3.3.1.2 Organización censal para la recolección.
La Unidad Operativa asumió la recolección de la infor-
mación. En cada regional del DANE se estableció una
estructura operativa y administrativa para apoyar el
operativo. En cada departamento se tuvo una organiza-
ción con un delegado y asistentes técnicos y adminis-
trativos. La organización departamental tuvo Jefes de
Zona diferentes a los del DANE, quienes asumieron la
supervisión y el control del censo en diez municipios,
en promedio. Los municipios también tuvieron delega-
dos y asistentes y se dividieron en dos grandes grupos:
menores de 3,000 viviendas y mayores de 3,000 viviendas.
Los municipios más grandes estaban divididos en comunas
y a cada una de ellas se le adjunicó un Jefe de Comuna.
Estas a su vez se dividieron en Centros de Operaciones
los cuales estuvieron a cargo de un Jefe de Centro y
dos auxiliares. A cada Centro de Operaciones le corres
pondieron 3,000 viviendas, en promedio. Para llegar a
esta cifra se hizo un análisis de sensibilidad donde se
tuvieron en cuenta los costos fijos y variables, el
área y el número de viviendas. Cada centro tuvo doce
empadronadores y tres jefes de grupo, en promedio y ca
da empadronador asumió una carga de trabajo por día de
unos 30 formularios básicos o 10 formularios ampliados.
Los municipios de tamaño mediano tuvieron una organiza
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
73
ción similar. Todos los municipios tenían coordinado-
res rurales con grupos de empadronadores encargados de
recoger el censo rural de cada municipio. El resto,
por ser más pequeños que un centro de operaciones, no
contaban con jefes de centro, sino que eran atendidos
directamente por los delegados; en este caso, las mis-
mas personas censaron la zona urbana y la rural.
Los Centros de Operaciones fueron claves en el desarro
lio del operativo. La definición del área de los cen-
tros de operaciones permitió que se pudiera estandari-
zar y controlar la carga de trabajo y todo lo relacio-
nado con los suministros y elementos de oficina. Una
de sus características importantes fue que nunca se
fraccionaron las divisiones estadísticas.
En los centros de operaciones se planeó minuciosamente
y con suficiente anticipación el trabajo de los empa-
dronadores. El personal se vinculó entre uno y dos me
ses antes de iniciar el CENSO 85. Las oficinas regio-
nales del CENSO 85 se encargaron de la selección del
personal del operativo, con excepción de los delegados
departamentales, quienes fueron entrevistados y selec-
cionados por el Jefe del DANE y el Director del CENSO
85. Asimismo, realizaron los contactos para obtener
las oficinas, recopilar la cartografía, etc. Las of i-
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
74
ciñas municipales tuvieron a su cargo la elaboración
de nóminas y plantillas para el pago del personal.
3.3.2 Actualización cartográfica y recuentos de -
vivienda.
Se establecieron como objetivos de la cartografía los
siguientes:
a) garantizar máxima cobertura
b) optimizar la asignación de los recursos
c) servir de marco para el diseño, selección y apli-
cación de muestras; y
d) servir de base para la presentación y análisis
de los resultados.
Estadísticamente, los municipios grandes se dividen,
en la parte urbana, en comunas. Las comunas, los mu-
nicipios medianos y pequeños y los centros poblados se
dividen en sectores, secciones y manzanas. Todas estas
divisiones tienen asignado un código, de manera que son
distinguibles. Las áreas con población dispersa tam-
bién se dividen en sectores y secciones.
Para la actualización cartográfica se hizo un inventa-
rio de la cartografía existente tanto en las oficinas
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
75
del Estado, como en entidades privadas. Entre septiem
bre de 1984 y mayo de 1985 se realizó un recuento pre-
liminar de viviendas y actualización cartográfica de
la zona urbana. Este recuento contó viviendas desde
el exterior, sin precisar todas las viviendas censales,
ya que solo se requerían datos gruesos para efectos de
planeación y diseño. Este trabajo se realizó en las
capitales con la participación de personal de las ofi-
cinas de planeación de los municipios y universidades.
La participación de la comunidad en esta etapa fue de-
cisiva para lograr su colaboración posterior. La ac-
tualización cartográfica y el recuento de viviendas de
los municipios menores de 20,000 habitantes, los hi-
cieron los Jefes de Zona del DANE; para el resto se
contrató personal directamente. Todo el trabajo estu-
vo supervisado por cartógrafos asignados a las regiona
les .
La actualización cartográfica que se hizo con el re-
cuento preliminar, se digitalizó en microcomputadores.
La digitalización automática de la cartografía se dise
fió para contener información cartográfica y datos de
los recuentos de vivienda, y así lograr una mejor co-
bertura, una mayor confiabilidad de los datos censales
una mejor planeación de los recursos y la oportunidad
en la obtención de los resultados. Se digitalizó el
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
92% de los municipios, equivalente al 75% de las rnanz
ñas y pasó por control de calidad el 84%, lo cual s.q
nificó cerca del 60% de las manzanas de las zonas urb
ñas de 1 pals.
A partir del 2 de septiembre los jefes de Centros de
Operaciones y sus auxiliares realizaron un recuento 1
nea a línea o precenso de las viviendas de su área.
Este recuento incluyó la dirección, el nombre del jet
de la vivienda y si era posible el número del jefe de
la vivienda y si era posible el número del teléfono.
Con este recuento se trataba de garantizar un mejor
control y una mayor cobertura. A pesar de la diligen
cia de los recolectores, no siempre se logró un acces
adecuado a las viviendas y durante el censo se descu-
brieron viviendas adicionales en algunos sitios. El
recuento precensal registró 10.6% más de viviendas qu
el recuento preliminar y el censo 2.9% más viviendas
que el precenso.
3.3.3 Formularios, pruebas piloto y censos experime
tales.
3.3.3.1 Formularios.
Se utilizaron cinco formularios:
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
77
1) básico
2) ampliado
3) básico de viviendas colectivas (en inglés y español),
4) indígena (uno en español y dos en idioma indígena), e
5) indígena de viviendas colectivas.
El básico tiene un conjunto de preguntas fundamentales
sobre las viviendas y las personas, ya que responde al
objetivo de obtener información necesaria para el marco
muestral. Las preguntas contenidas en el básico se en-
cuentran también en el ampliado. Este completa la in-
formación censal, por medio de una muestra, con pregun-
tas demográficas, laborales y de educación. El formu-
lario indígena se diseñó especialmente para esa pobla-
ción . Asimismo, se diseñaron formularios para perso-
nas que viven en viviendas colectivas, tanto indígenas,
como no indígenas.
Los formularios se diseñaron de manera que satisfacie-
ran varios requisitos:
a) que contuvieran las preguntas estrictamente necesa-
rias para el cumplimiento de los objetivos del CEN-
SO 85. Por lo tanto, no se incluyeron variables
para las cuales el censo no es el instrumento ade-
cuado para medirlas.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
78
b) que fueran simples, lo cual facilitaría los proce
sos de entrenamiento, inteligibilidad y diligencia
miento, para lograr eficiencia y calidad; y
c) que garantizaran la oportunidad en el procesamien
to de los datos y la producción de los tabulados.
Para obtener información confiable y oportuna, se adop
tó la decisión inflexible de cerrar todas las pregun-
tas con excepción de unas pocas en los formularios in-
dígenas. Asimismo, los formularios se diseñaron bus-
cando la mayor legibilidad posible. También se trató
de minimizar el consumo de papel y la producción de des-
perdicio y, a la vez, maximizar el número de personas
incluidas para evitar la utilización de formularios
adicionales. En el básico tuvieron cabida 21 personas,
en el ampliado 14 y en el básico para viviendas colec-
tivas , 63 .
El Comité Técnico sugirió un formulario según esos 1i-
neamientos. Posteriormente se le introdujeron cambios
de acuerdo con las pruebas de contenido realizadas en
julio de 1984. En octubre de 1984 se realizó un semi-
nario con miembros del Comité y otros usuarios para
discutir el contenido, la redacción de las preguntas y
ios tabulados más recomendables.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
79
Del seminario con los usuarios resultó un formulario que
fue discutido ampliamente con el equipo del CELADE,el
cual desempeñó un papel decisivo en la versión final.
Se aclara en forma enfática que el aporte del grupo de
CELADE fue significativo, pero, la responsabilidad del
contenido final fue del Director del CENSO 85 y del Je-
fe del DANE. Esto se hace explícito porque el grupo
del CELADE no comparte parcialmente la versión final.
Esta discrepancia se dio básicamente en el cambio de la
unidad de análisis -viviendas y no hogares-, la elimina
ción de preguntas abiertas sobre rama de actividad de la
empresa donde trabaja la persona, su ocupación y migra-
ción de todos los municipios del país.
Para definir el formulario indígena, se creó un grupo
formado por el DANE y entidades oficiales y privadas.
Desde principios de 1985, se tuvieron reuniones con or-
ganizaciones indígenas nacionales y se consultó a las
organizaciones indígenas regionales y locales.
3.3.3.2 Pruebas piloto.
En julio de 1984 se llevaron a cabo pruebas de conteni-
do y redacción de las preguntas de los formularios bási
co y ampliado en las zonas urbanas y rurales de 15 muni
cipios situados en 13 departamentos y una comisaría. Es
tas se realizaron con personal experimentado en esas ac
tividades para no introducir fallas adicionales.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
80
3.3.3.3 Censos experimentales.
Los objetivos de los censos experimentales o minicen-
sos, fueron garantizar que en la planificación se con-
templaran todas las actividades y que los soportes lo-
gísticos y administrativos fueran adecuados v ,en gene
ral, evaluar en forma integral los resultados obteni-
dos. En los minicensos se sometieron a prueba la ma-
yor cantidad de aspectos, en particular, la ejecución
del trabajo administrativo y operativo.
Se realizaron 6 minicensos en 5 departamentos. En no-
viembre de 1984 se realizaron tres en municipios menores
de 10,000 habitantes. En dos de ellos se realizó censo
urbano únicamente y en el otro se hizo censo urbano y ru
ral. En marzo de 1985 se hicieron dos censos urbanos
experimentales: uno en una ciudad de casi 180,000 habi-
tantes y otro en una de casi 50,000. El primero lo eje
cuto y procesó una universidad local bajo la supervisión
y asesoría del CENSO 85. El segundo se hizo con perso-
nal propio y se tuvo la oportunidad de probar todas las
etapas, incluido el procesamiento de los datos. En ma-
yo de 1985 se realizó otro en la zona rural de un muni
cipio de 20,000 habitantes. Los censos urbanos se rea
lizaron de hecho y los rurales de derecho.
Uno de los resultados de los minicensos fue encontrar
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
81
que el concepto de hogar que se ha manejado en los
censos, presentaba serias dificultades de conceptúa™
lizaciÓn. Ello condujo a cambiar el concepto de hogar
tradicional, por el de "grupo de personas que compar-
ten una vivienda". Esto significó que la unidad de
análisis es la vivienda y no el hogar. Asimismo, al
examinar las cifras del censo de 1973 se concluyó que
la tipología tradicional de vivienda no era adecuada,
porque mezcla aspectos cuantitativos con aspectos cua-
litativos. La tipología adoptada con el tipo de ma-
terial con que están construidos los pisos y las pare-
des .
Aunque estas pruebas fueron exitosas, precisamente sus
resultados hicieron refelexionar sobre la probabilidad
de obtener un éxito similar en forma simultánea en mil
municipios. Esto y el costo social en que el país in-
curría por la paralización, calculado en más de
US $40 millones, condujeron a adoptar la decisión de
censar a la población de derecho sin inmovilización,
durante un período de quince días. Esta estrategia
también se adoptó porque contribuía a garantizar una
alta cobertura, la oportunidad y confiabilidad de los
resultados y a disminuir la probabilidad de falle en
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
82
el operativo.
3.3.4. Infraestructura para la recolección.
Garantizar los recursos necesarios para una buena re-
colección es condición básica para hacer un censo exi
toso. Entre los puntos más importantes por conside-
rar en ese aspecto están:
a) materiales de empadronamiento,
b) comunicaciones,
c) distribución y recolección de materiales,
d) reclutamiento y selección de personal, e
e) instalaciones físicas y dotación.
3.3.4.1 Materiales de empadronamiento.
La experiencia de 197 3 y las recomendaciones de los
asesores indicaron que se debían producir cantidades
de material con excedentes suficientes. Se imprimió
37% más de formularios básicos y 52% más de ampliado,
sobre la base del recuento preliminar de viviendas.
En cuento a los materiales de capacitación, se impri-
mió una cantidad adicional menor. Estos excedentes
se repartieron a todos los niveles de la organización
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
83
de manera que se pudieron satisfacer las necesidades
que se presentaron en algunos sitios.
La impresión de descentralizó para reducir las conse-
cuencias de posibles incumplimientos de los impresores.
Este trabajo se hizo en las seis ciudades donde el
DANE tiene regionales. Se contrató con empresas pri-
vadas y se mantuvo la imprenta del DANE para emergen-
cias y necesidades menores.
3.3.4.2 Comunicaciones.
Para facilitar la solución de los problemas y contar
con una información preliminar oportuna se estableció
una red nacional de comunicaciones. Se abrió una
cuenta corriente para que los funcionarios encargados
del CENSO 85 se comunicaran con las delegaciones de-
partamentales y éstas con las regionales. Todas las
oficinas departamentales contaron con télex o tenían
acceso libre a él en alguna entidad estatal. Con ello
se logró mantener comunicación permanente sobre reclu
tamiento, contratación, capacitación, recepción de ma
teriales, recolección de información, pagos, etc. Es
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
84
ta red permitió conocer a partir del mismo 15 de
octubre, los datos preliminares de muchos municipios.
3.3.4.3 Distribución y recolección de materiales.
En los minicensos se probó la capacidad de distribu-
ción. Se mostró que el "negocio" del DANE es la
producción de información y no el transporte, por lo
tanto, lo más razonable era contratar ese servicio
con firmas especializadas.
El diseño de los centros de operación permitió empa-
car y distribuir el material en forma más eficiente.
Este llegó a las bodegas de la firma y, con instruc-
ciones precisas, procedió entregarlo a los centros
por lo menos una semana antes de iniciarse el CENSO 85.
Los jefes de centro tenían instrucciones precisas para
la recepción, el empaque y la entrega del material. Ca
da caja fue cerrada con un sello de seguridad, para evi
tar alteraciones y garantizar la confidencialidad de la
información. Los formularios diligenciados y el mate
rial sobrante, fue recogido por la firma y entregado
en los sitios donde se realizó la grabación.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
85
3.3.4.4 Reclutamiento y selección de personal.
Siempre se mantuvieron dos criterios claves: mantener
el CENSO 85 alejado de los intereses electorales y
utilizar la infraestructura del Estado siempre y cuan-
do garantizara la eficiencia del proceso. El CENSO 85
fue un botín apeticido por algunos políticos, por lo
tanto, se diseñaron estrategias para mantener el proce
so ele jado de influencias indeseables. El reclutamien
to de personal fue abierto y a través de diversos cana
les: el Servicio Nacional de Empleo (SENALDE), las al-
caldías municipales y la recepción directa de solicitu-
des en todas las oficinas del CENSO 85. Este proceso
liberó al grupo encargado del proyecto de las presiones
de los solicitantes.
SENALDE y las alcaldías publicaron avisos de prensa y en
la radio para convocar a los interesados. En los munici_
pios pequeños esta labor se hizo de vida voz o por al-
tavoces en los sitios más concurridos. En las capitales,
el personal reclutado debió aprobar pruebas de aptitud,
administradas por el Servicio Nacional de Aprendizaje
(SENA), o comprobar haber obtenido un mínimo puntaje en
los exámenes nacionales de final de estudios de bachi-
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
86
llerato. Además, todo el personal debió aprobar las
pruebas de evaluación de los cursos de capacitación. Es
tas pruebas fueron el instrumento final para la selec-
ción .
Los cursos se hicieron con manuales, conferencias, ejer-
cicios en el aula y en el campo. Los capacitadores fue
ron funcionarios del CENSO 85, quienes a su vez capacita
ron a otros. Como norma general, éstos fueron profesio
nales con experiencia docente o estudiantes de último año
de Ciencias de la Educación.
Se reclutaron 90,000 personas en total, de las cuales
31,000 se recibieron a través de SENALDE, 39,679 se some
tieron a pruebas de aptitud del SENA; se capacitaron
50,573 y se seleccionaron 40,000.
3.3.4.5 Oficinas y recursos físicos.
En cada oficina regional, departamental, municipal y zo-
nal se contó con instalaciones físicas adecuadas para el
desarrollo del trabajo. Entidades oficiales y privadas
prestaron las oficinas y sus dotaciones. En dos depar-
tamentos se suscribieron convenios con universidades pa
ra que coordinaran el censo y, además, aportaran las ins
talaciones físicas. Se consiguieron en forma gratuita
más de 2000 oficinas para las actividades de recolección,
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
87
de grabación de da:os y de digitalización. De la misma
forma se obtuvieron vehículos para el personal de super
vision.
3.3.5 Publicidad.
Se contrató una campaña publicitaria con la más importan
te firma del ramo, después de haber realizado un concur-
so de mentes con cerca de una decena de empresas. Al
comienzo se consideró conveniente contratar a una firma
pequeña, pero ésto fue rectificado y se asignó el contra
to a una firma que además de calidad, ofreciera un amplio
respaldo financiero. Esto fue acertado porque los t rámi^
tes burocráticos demoraban la liquidación de los pagos;
por lo tanto, la firma debía financiar la campaña por tiem
por cortos, como en efecto sucedió.
La campaña se dirigió a toda la población y transmitió
una imagen ágil, eficiente y amable del censo, sin conno
tac ion de proyecto estatal. Hizo hincapié en un censo
diferente, con tecnología muy avanzada, sencillo y que,
además, no paralizaría al país. Se insistió en la im-
portancia de los datos para el desarrollo del país y en
la confidencialidad de la información. Se inició el 7
de julio de 1985 y tuvo cuatro etapas: lanzamiento, sos-
tenimiento, censo y divulgación de resultados. Cada eta
pa respondió a un objetivo específico del proyecto y a
un objetivo de comunicación.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
88
El primer día, se transmitió por televisión una entre-
vista que hizo uno de los mejores periodistas del país
al Jefe del DANE, e inmediatamente se lanzaron al aire
las cuñas de televisión y radio. El día siguiente apa
recieron avisos a color y en blanco y negro en todos
los periódicos del país. Posteriormente, se publicó
un aviso con el texto de un plegable explicativo sobre
el censo. Cerca del 15 de octubre se publicaron avisos
que mostraban los formularios básicos y ampliado. Se ins
talaron vallas en las principales ciudades, se distribu-
yeron 100,000 afiches, se repartieron plegables y un fo-
lleto de 8 páginas, claro y sencillo, sobre cómo se rea-
lizaría el CENSO 85. También se emitió una estampilla
de correo con el símbolo del CENSO 85. El 15 de octu-
bre se presentó un programa de televisión, para explicar
cómo se haría el CENSO 85.
Se enviaron miles de cartas autógrafas del Director del
CENSO 85 a los medios de comunicación, al clero, a los
alcaldes, a las empresas públicas y privadas, y a las
universidades solicitando apoyo. Asimismo, se hicie-
ron numerosas conferencias, entrevistas y ruedas de pren
sa. Además, la Oficina de Prensa mantuvo informados a
los medios de comunicación sobre los avances del proyec
to.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
89
El clero desempeñó un papel decisivo. Se envió un afi^
che y una carta firmada por el director de una entidad
que agrupa a todos los párrocos del país, donde se sol_i
citaba su apoyo al CENSO 85. Esta estrategia es muy
valiosa, sobre todo en países donde la influencia de la
Iglesia Católica es muy fuerte.
El 22 de diciembre se publicó un aviso de una página,si-
milar a los anteriores, en el cual se agradeció a todos
los habitantes por su colaboración. Se envió un folle-
to de los resultados preliminares a los gobernadores y
alcaldes, con una carta autógrafa del Director del CENSO
85, agradeciéndole a él y a toda la población su colabo-
ración .
La División de Divulgación logró una colaboración sin
precedentes: se obtuvo de los medios de comunicación des
cuentos entre el 35% y el 93.4% y, en algunos casos, pu-
blicaron avisos de prensa y mensajes de radio y televi-
sión sin costo; en los teatros se presentó la cuña de
televisión en la misma forma. En general, se logró un
apoyo incondicional de la comunidad. Todo esto se re-
flejó en que el precio de la campaña fue muy inferior a
su valor comercial: costó 470 mil dólares y su valor co-
mercial se calculó en 3.7 millones de dólares.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
90
3.4 Eventos Censales.
Estos se pueden relacionar, en forma secuencial, así:
a) Censo urbano,
b) censo rural e indígena,
c) muestra de ampliación, y
d) muestra de cobertura.
3.4.1 Censo urbano.
El censo se realizó sin inmovilización aproximadamente
en quince días y registró a la población de derecho el
día 15 de octubre, tanto en viviendas particulares, co-
mo en colectivas. Por razones de clima, topografía y
transporte, unos pocos municipios se censaron durante
el primer trimestre de 1986, pero siempre con el 15 de
octubre de 1985 como momento censal.
La recolección se realizó en forma tradicional, excepto
en conjuntos habitacionales cerrados, que contaran con
algún tipo de organización comunitaria. Allí se reali
zó el autoempadronamiento dirigido, que consistió en
que los formularios se dejaban para que las personas lo
diligenciaran y el empadronador lo recogía con posterio
ridad. Esta modalidad se utilizó sólo con el formula-
rio básico.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
91
Para aumentar la cobertura, en las grandes ciudades se
instalaron puestos de información en sitios muy concurrí
dos, donde las personas que no hablan sido censadas de-
jaban su dirección para que las censaran.
3.4.2 Censo Rural e indígena.
A partir del 22 de octubre se censaron las zonas rura-
les de los departamentos. En este caso la duración
prevista fue de un mes. Como en el censo urbano, hubo
regiones que se censaron durante el primer trimestre de
1986. Las zonas rurales de las intendencias y comisa-
rías tuvieron un tratamiento especial, debido a que en
su mayoría son regiones selváticas y de llanura con po-
blación muy dispersa. Se censaron durante el primer
trimestre de 1986 para aprovechar las condiciones clima
ticas más propicias.
La población indígena que habita en resguardos,reservas,
o áreas de asentamiento indígena, fue censada con un for
mulario especial, adaptado a sus características cultu-
rales. El censo fue simultáneo con el censo rural. Los
empadronadores fueron miembros de las comunidades indige
ñas que contaban con su aceptación y bilingües.
3.4.3 Muestra de ampliación.
La muestra fue cocensal, sistemática del 10% con arran-
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
92
que aleatorio. El formulario ampliado se diseñó para
recoger la muestra y el formulario básico se aplicó al
resto. Con esta estrategia se trató de obtener infor
mación más confiable y más oportuna, al no tener que
procesar grandes cantidades de formularios con demasía
das preguntas. Se aplicó en forma simultánea en zonas
urbanas y rurales, exceptuadas las áreas indígenas. La
precisión esperada es aceptable para municipios mayores
de 16 , 000 habitantes, en el caso de la variable más crí.
tica.
En la zona urbana fue fácil asignar y controlar la niucs
tra, con los recuentos de viviendas línea a línea. Pa-
ra evitar que los empadronadores la sesgaran, aplicando
el formulario a viviendas con pocos habitantes se asig-
nó, en los centros de operaciones, especificando la di-
rección, el sitio de la encuesta. Además, en los muni
cipios mayores de 3,000 viviendas, quienes recogieron la
muestra fueron diferentes a quienes aplicaron el formula
rio básico.
3.4.4 Muestra de cobertura.
Para medir la cobertura se diseñó una muestra con más de
28,000 encuestas. Se asignaron 1,000 a cada departamen
to, 2,000 a Bogotá D.E., 1,000 a Cali y Medellin y 1,000
a las intendencias y comisarías, como un todo. El mués-
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
93
treo se hizo por áreas, con segmentos promedio de 25 vi^
viendas y un máximo de 40. La muestra se recogió en
zonas urbanas y rurales.
Por medio de algunas preguntas y un proceso de apareamien
to con los formularios censales, se midió la cobertura en
dos formas: por medio de la percepción del entrevistado y
por medio de la comparación de los formularios. La mués
tra de cobertura urbana se recogió entre el 7 y el 22 de
diciembre de 1985; la rural entre el 10 de febrero y el
10 de marzo. Los datos preliminares de la muestra de co-
bertura y en particular la recordación de las personas
respecto de si las viviendas habían sido censadas o no,
indican una cobertura de 96.8% en lo urbano y 93.1% en lo
rural, lo cual en promedio, indicaría una cobertura de
95.6%. La cobertura calculada con base en las proyeccio
nes disponibles está entre 95.8% y 98.7%.
Ambas muestras fueron asesoradas por la Oficina del Censo
de los Estados Unidos. Para garantizar mayor independen
cia de la muestra, el personal que la recolectó, no par-
ticipó ni en la supervisión, ni en la recolección censal.
3.5 Eventos Postcensales.
Recogido el censo, se iniciaron las actividades postcensa
les asi:
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
94
a) grabación de datos censales,
b) evaluación, procesamiento de datos y producción de
tabulados, y
c) publicación de resultados.
3.5.1 Grabación de datos censales.
Inicialmente se consideró la lectora óptica para la gra
bación, pero se descartó con base en la experiencia de
algunos países y otras consideraciones. Las principa-
les razones para eliminarla, además de esa experiencia,
fueron: las condiciones físicas en que debían mantener-
se los formularios (sin dobleces, completamente secos,
etc.), la subutilización del equipo una vez terminado el
censo y,por último, la calidad de papel y de impresión
requeridas, lo que habría hecho necesario importar los
formularios. Ninguna de estas situaciones le convenían
al país, en lo económico, ni al censo, en político. Tam-
bién se alcanzó a considerar la posibilidad de imprimir
formularios con respuestas codificadas en barras y gra-
barlas directamente en el campo, pero también se desear
tó. Finalmente, se optó por hacer la grabación en mi-
crocomputadores.
La empresa
gradual en
6 ciudades
transportadora entregó el material en forma
ocho centros de captura de datos ubicados en
El primer paso allí es el almacenamiento
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
95
en archivo físico. Este material, rotulado y empacado
en bolsas de plástico y en cajas de 120 formularios bá
sicos o 30 formularios ampliados, se organizó en están
terías. Las cajas y las bolsas tienen una identifica
ción completa del municipio, del centro de operaciones,
del sector, de la sección, de la manzana y casillas de
control para los diferentes procesos. Allí se criti-
caron, revisaron y numeraron. El proceso de crítica
revisó la identificación geográfica de los formularios
y que el total de personas anotadas correspondiera con
el número de personas realmente censadas. Este proce
so tuvo un control de calidad que rechazaba o aceptaba,
por muestreo, lotes de formularios, según el porcenta-
je de errores; los lotes rechazados fueron criticados
nuevamente. Una vez criticados, los formularios se nu
meraron; esta numeración fue su identificación definite
va.
Los formularios se grabaron con programas diseñados por
la Unidad de Procesamiento de Datos del CENSO 85. Los
programas de control de materiales, grabación y verifi-
cación, fueron desarrollados por una firma especializa-
da. La grabación se contrató con dos firmas expertas
en captura inteligente de datos. El resto de los proce
sos fueron asumidos por personal del CENSO 85. Las per
sonas que hicieron la grabación se sometieron a un entre
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
96
namiento en transcripción de datos y, específicamente,
en la de los formularios censales. Para estas labores
se contrataron más de 1,200 personas, incluidos los fun
clonarlos de las firmas contratistas. La descentrali-
zación y la contratación con firmas especializadas en
grabación de datos permitieron una mayor rapidez en ese
proceso y contribuyeron a la oportunidad de los resulta
dos. Además, la descentralización coadyuvó a que el
CENSO 85 tuviera amplia aceptación en la comunidad.
La verificación tuvo como objetivo garantizar la fideli
dad de la grabación. Los formularios se seleccionaron
por muestra y se grabaron de nuevo para su verificación.
El programa revisa que los datos coincidan con los gra-
bados y en caso contrario, se verifica si se trata de
un error en uno u otro proceso. En caso de que el núme
ro de errores detectados pase de cierto nivel, se recha-
za toda la caja de donde procede el formulario y debe
ser grabada de nuevo.
El control de materiales mantiene información permanente
sobre el material. Con los datos preliminares obteni-
dos de cada centro de operaciones, se estableció el nú-
mero de formularios esperados en cada centro de captura,
y con ello se realizó un control primario de cobertura,
por medio del cual se sabía qué material ha llegado y
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
97
cuál falta por llegar. El programa identificó las ca-
jas y los "diskettes". Este proceso asignó el traba-
jo a cada digitador, permitió conocer el estado de ca-
da caja y decidió cuándo se debía criticar, grabar, ve
rificar o transmitir. Una vez terminado, cada proce-
so actualiza el estado del material. Cada día informa
sobre rendimientos en cada proceso, novedades del per-
sonal, estadísticas de errores y de producción por su-
pervisor, ausencias del personal, etc. Con este progra
ma se transmiten los datos de "diskettes" a una "casse-
tte" o cartucho, y de allí se pasa la información al
computador central.
Para probar y ajustar todos los procedimientos se insta
ló un centro de captura piloto desde diciembre 9 de 1985
hasta enero 11 de 1986. Aunque en esta prueba se detec
taron errores y se hicieron modificaciones a los progra
mas, sobre la marcha, con otros centros de captura ins-
talados, hubo necesidad de hacer nuevos ajustes. En es-
te centro se estudiaron las curvas de aprendizaje de ca
da proceso y se establecieron los parámetros finales de
trabajo. Los otros centros iniciaron sus labores en for
ma sucesiva de manera que la experiencia acumulada se
aprovechó en cada caso. El primero inició actividades
el 7 de enero de 1986 y el último el 28 de febrero del
mismo año.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
98
Para controlar el avance de la grabación se producía un
informe diario de lo realizado en los centros de captu-
ra de todo el país, en el cual se indicaba la cantidad
de formularios que se trabajó y el acumulado, se indica
ba el promedio de formularios a la fecha y los días que
faltarían para terminar, si los rendimientos se mantu-
vieran iguales a los de ese día.
Para garantizar la publicación oportuna de los resulta-
dos se adquirieron 170 microcomputadores, de los cuales,
30 se utilizaron para la digitalización cartográfica.
3.5.2 Evaluación, procesamiento de datos y tabulados.
3.5.2.1 Evaluación.
El grupo de evaluación tuvo como objetivo garantizar la
consistencia interna de los datos resultantes y la def i_
nición y revisión de los tabulados para publicación.
También tuvo como tarea elaborar el documento metodoló-
gico en donde se compara lo diseñado con lo ejecutado en
operativo. A este grupo se vincularon instituciones
que han tenido amplia experiencia en esta clase de tra-
bajo y que tenían un interés especial en participar en
el proceso.
Tuvo a su cargo la asignación de inconsistencias con el
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
99
programa CONCOR de la Oficina del Censo de los Estados
Unidos y con la asesoría del CELADE. La primera versión
de los planes de inconsistencias se utilizó en los cen-
sos experimentales y sirvió de base para la versión defi
nitiva que se implemento. Se adoptó una política de mo
dificar lo menos posible los datos, de manera que el
usuario recibiera la información lo más pura posible.
Como complemento de la prueba de cobertura, el grupo ana
lizó los datos del censo de 1973, y de encuestas cerca-
nas a la fecha del censo. Asimismo, se utilizaron los
resultados de los minicensos. Esta información sirvió
como punto de referencia para comparar los datos del CENSO
85 con censos anteriores y para analizar y observar las
tendencias encontradas.
3.5.2.2 Procesamiento de los datos.
El procesamiento de los datos lo hizo personal del CENSO
85 en un computador IBM 4361. Una vez generadas las cin
tas con los cartuchos de los centros de captura, se gene
raron los lotes a procesar y se realizó el control prima
rio. Inmediatamente se procedió a la corrección de in-
consistencias, La expansión y la producción de tabulados.
Como el procesamiento se inició antes de finalizar la gra
bación de los datos, se pudieron detectar algunos proble
mas en los programas de grabación y transmisión. Sin em
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
100
bargo, el procesamiento se inició casi al final de la
grabación de los datos, con los riesgos que ello implica
3.5.2.3 Producción de tabulados.
Se definieron 16 tabulados para publicación. Estos tabu
lados fueron producidos por medio de un paquete CENTS-3,
también de la Oficina del Censo de los Estados Unidos y
con asesoría del CELADE.
3.5.3 Publicación de resultados.
El día 20 de diciembre de 1985 se entregó al país un fo-
lleto con los resultados preliminares del CENSO 85, co-
rrespondientes al 93% de la población, y en enero de
1986 se publicó otra edición con información sobre el
96.4%. A estos resultados fueron agregados manuales rea
lizados por el personal de campo en todos los niveles de
organización.
Aunque se publicaran pocos tabulados, ésta no es la úni-
ca manera de tener acceso a los datos definitivos del
CENSO 85. Estarán disponibles en volúmenes impresos,
en un folleto-resumen insertado en un periódico de alta
circulación nacional, en microfichas, "diskettes" y cin-
tas. A mediados de julio de 1986 se publicarán seis vo-
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
101
lúmenes con los datos definitivos. El primero contiene
la metodología del CENSO 85 y los datos generales de la
población por sexo (para departamentos y municipios, in-
cluyendo población indígena) y por sexo y edades para ca
da departamento y su capital, además, se incluye informa
ción en forma gráfica. El segundo volumen contiene da-
tos sobre educación y empleo. Un tercer volumen muestra
los datos de variables demográficas: migración, fecundi-
dad y mortalidad. Las variables sobre vivienda ocupan
un cuarto volumen. El quinto volumen registra a la pobla
ción indígena, con tabulados similares a los del resto de
la población. El sexto volumen contiene información de
los censos anteriores, ajustados a la división político-
administrativa vigente.
Una de las dificultades encontradas en el diseño del CENSO
85 fue la carencia de información sistematizada de la ex-
periencia de 1973. Para evitar eso en el futuro se contra
tó con una universidad la elaboración de la memoria del
proyecto. Allí se registran los avances y retrocesos que
se tuvieron durante el proyecto, las alternativas estudia
das y las razones que llevaron a adoptar la seleccionada.
Se decidió contratar una entidad externa, para evitar que
las relaciones de autoridad y la misma dinámica del pro-
yecto impidieran una visión objetiva de lo que sucedía.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
102
4. MANEJO ADMINISTRATIVO Y FINANCIERO.
4.1 Administración.
Los fondos que recibe el DANE para sus proyectos provie-
nen del presupuesto nacional y los maneja a través del
Fondo Rotatorio del DANE (FONDANE). Una vez apropiados,
se debe realizar un gran número de trámites burocráticos
que pueden durar, en condiciones normales, más de dos me
ses. Los pagos de operaciones aprobadas, pueden durar
entre 15 y 20 días hábiles. Uno de los méritos del equi_
po encargado del manejo administrativo del CENSO 85 fue
haber sido eficientes a pesar de la lentitud de la admi-
nistración pública.
4.2 Costos.
Desde 1983 hasta 1986, el gobierno asignó al CENSO 85 un
presupuesto de US $18,447,664 y se ejecutó de la siguien
te forma:
AÑO US $
1983 87,394
1984 2,080,918
1985 11,597,415
1986 2,199,438
Total 15,965,165
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
103
Esto significa que en el censo se gastó el 86.54% de los
fondos asignados. A esta suma debe agregarse el apoyo
de las Naciones Unidas por un monto de 407 mil dólares.
Para comparar el costo de un censo con el de otros paí-
ses , debe calcularse el costo percápita. El número de
habitantes censados en Colombia es de 27.7 millones, por
lo tanto, el costo percápita es de menos de 60 centavos
de dólar.
5. EVALUACION DE LA GESTION.
Con la seguridad de que los lectores podrán anotar otros, a
continuación se presentan los aspectos negativos y positivos
que se consideran más importantes.
Entre los aspectos negativos se deben mencionar:
La legislación demasiado restrictiva impidió una mayor
eficiencia. Esto fue la causa de muchos de los problemas
y de lo que aquí se presenta como aspectos negativos de
la gestión.
La baja calidad del informante de los censos sin inmovili
zación se agravó por la sencillez y corta extensión del
formulario básico. Esto pudo generar problemas de medi-
ción en variables como la edad. Además, ello produjo en
parte de la población la sensación de no haber sido censa
da.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
104
Se exageró la sofisticación de procesos y controles.
Por ejemplo, el programa de control de materiales utili-
zado en los centros de captura es excelente para un pro-
ceso mucho más largo: el tiempo de ajustes fué casi igual
al de su uso. Hubo exceso de formas de control del opera-
tivo lo cual le restó importancia a algunas formas claves
que en algunos casos no se diligenciaron o no se recibie-
ron.
Se contó con muy poco tiempo para preparar y ejecutar el
censo. Esto se agravó con el cambio de metodología, que
se hizo seis meses antes de iniciar el censo.
El centro de captura piloto debió iniciar operaciones con
más anticipación y durante un tiempo más largo.
El capturador central se instaló tardíamente y se cambió
de sistema operacional y de proveedor a última hora.
No se hizo un adecuado traslado entre el procesamiento y
la grabación de los datos, que permitiera detectar pro-
blemas en la grabación de la información.
De los seis censos experimentales sólo se evaluaron rigu-
rosa y exhaustivamente tres de ellos: dos de municipios grandes y
uno pequeño. En los demás solo se evaluó el aspecto operativo.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
105
Se subestimó la magnitud de los procesos administrativos
en la organización para la recolección.
Alta rotación de personal de dirección: Entre agosto de
1984 y finales de 1985 se retiraron 9 jefes de 5 unidades.
De estos retiros, seis fueron voluntarios; la unidad de
procesamiento de datos tuvo cuatro jefes.
La descentralización que se le dió al CENSO 85 fue un ex-
perimento, en el sentido que nunca antes las regionales
hablan tenido esa autonomía y no se les entrenó lo sufi-
ciente para ello.
Entre los aspectos positivos se deben mencionar:
Organización no burocrática y separada del DANE, con obje
tivos claros y precisos. Dentro del grupo las comunica-
ciones internas fueron muy informales y ágiles. Se logró
crear una organización dinámica capaz de adaptarse a cam-
bios organizacionales bruscos y sin ser consultados.
Existió un sano "laissez-faire" en el grupo de manera que
se mantuvo en algo la motivación y la mística. Todas las
personas del equipo encargado del CENSO 85 tenían acceso
fácil y directo a la Dirección.
Administración por resultados, lo cual fue un incentivo
para atraer a profesionales muy capacitados y con alta ne-
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
106
cesidad de logro. El trabajo se hizo por objetivos.
El grupo no tenía experiencia en manejo de censos, lo cual
propició el desarrollo de ideas imaginativas y no conven-
cionales .
Las trabas de tipo legal fu
tro de las normas legales e
se hubiera trabajado con un
tal, aquellas hubieran sido
ron superadas con ingenio den
istentes. Es posible que si
grupo de la burucracia esta-
res trice iones insalvables.
Se identificó con claridad el "negocio" del DANE, lo cual
condujo que actividades críticas se subcontrataran con em
presas privadas o, en general se aprovechara la infraes-
tructura existente en entidades públicas y privadas. Lo
anterior y el manejo del proyecto como una empresa priva-
da fueron mecanismos que permitieron superar las ineficien
cías impuestas por el marco legal y lograr una eficiencia
muy alta. En general, estas ideas se pueden considerar
positivas.
Los centros de operaciones fueron decisivos para la planea
ción y el control del operativo. El trabajo previo reali.
zado por los jefes y sus auxiliares sirvió de entrenamien
to a los mismos para conocer el terreno y, por lo tanto,
controlar mejor el censo. La existencia física de sitios
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
107
para la supervisión del trabajo de campo permitió una me-
jor planeación y un mejor control de la calidad y del ren
dimiento. La creación de los centros de operaciones con
la suficiente anticipación, permitió una buena planeación
del operativo lo más cercana al sitio de trabajo.
Se crearon instrumentos de planeación muy buenos. De he-
cho, los excedentes asignados a la producción del material
no fueron utilizados en la mayoría de los casos. Estos
márgenes relativamente algos aseguraron el suministro ade
cuados en todos los niveles de la organización.
Se trató de minimizar el riesgo y no los costos; a la lar
ga, lo primero conduce a reducirlos.
El sistema de archivo y de c
centros de captura, aunque s
considerarse excelente y su
carse en otros países.
ontrol de materiales en los
ofisticado en exceso, puede
filosofía general pueda api i -
En la sede del CENSO 85 se asignó una persona a cada re-
gional para atender exclusivamente sus problemas durante
el operativo. Esto permitió resolverlos mejor y con más
oportunidad.
El uso de microcomputadores para la planeación y la admi-
nistración facilitó el trabajo. Se utilizaron programas
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
108
pira p.] aiiíjac i ón y apoyo admia istrat i vo , éstos últimos no -
sarro] .1 ados por peí serial propio. Pai a e l uso de estos
p r oq r a ni. i s c a da a n i da ^ 1 t u v o di spo n i b 1 e po r 1 o me n o s un equi
po y se dispuso de un grupo para entrenamiento y solución
de consultas.
Iniciar el t i ana jo en todos Jos frentes, .incluidas i a s ül
tuna:; ' tupas como procesamiento de datos , fue una ventaja
para contaí oportunamente con instrumentos como los pla-
ne s de cwnsisienclas.
El censo s i a. i nmobi L i zación permit i ó una mejor atención a
ios p t < >b a. mu(a - pus i b I emerite una inay< u cobertura .
La a : :;cia' ¡ o¡¡ aobre ; a:. a taaavas de' ! ormu i ar .i.o censa 1 se
cono Luyo an di c.w,-ii¡bi. e 2U e 1 984, lo cual permitió avanzar
en ot ros p¡ ocasos. !■: I , d :o de me i ocio 1 oq í a y di 1 concep
to ai-: ;. a ¡; i • a ; . a . - as eecaLi icacioiios que se lueie
rer Ua— ¡ , n, [■ e-
c i .; n ¡ i ü;ífH i y ill tó lo:-, ¡ í aba
]OS Uü me i or cuma 1 inn en-
La campana p ubi ;a-t a i. i a l ue e x c e 1 e:, 1 e
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
109
El reclutamiento y la selección del personal permitió man
tener la imparcialidad que se necesita en el sector públi^
co para estos casos. Ni el Jefe del DANE, ni el Director
del CENSO 85 estaban desempeñando sus cargos a nombre de
ningún grupo político, lo cual permitió actuar con mayor
independencia.
6. RECOMENDACIONES Y CONCLUSIONES.
Aparte de lo obvio que es recomendar hacer lo positivo y evi-
tar lo negativo, conviene destacar algunas recomendaciones es
peciales.
Las personas que hicieron el CENSO 85 no deberían participar
en los próximos, sino como asesores. El hecho de no tener ex
periencia previa en estos proyectos puede permitir el aborda-
je de los problemas con mucha imaginación, se analizan más al^
ternativas y por lo tanto puede llegarse a mejores soluciones
Sin lugar a dudas, no se vacila en recomendar hacer censo sin
inmovilización y con muestra. Los países de América Latina
no se pueden dar el lujo de paralizar al país, aunque sea par
cialmente y además asumir el riesgo que implica hacer el cen-
so en un solo día y con formularios extensos.
Definitivamente, aunque en el CENSO 85 se tuvieron todas las
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
110
justifjcae iones del caso y se puede decir que fue acertado ha
cerlo, no se debe cambiar la metodología censal dentro del
año anteeior a su realj2ación; en la misma forma se debe pro-
ceder con las preguntas del formulario.
Pocas veces se ha demostrado en Colombia que se puede ser efi
cíente dentro del Estado y el CENSO 85 es una de ellas. Aun-
que los expertos en administración pública consideran que pre
cisamente esto fue una demostración de la bondad de la estruc
tura que enmarca las acciones oficiales y que lo determinante
en la ineficiencia estatal son las personas que aplican las
normas, no se vacila en recomendar que para cada censo se ex-
pida una ley eximiendo al proyecto de todos los trámites buró
oráticos que se exigen normalmente, o por lo menos, hacerlos
más flexibles.
El prototipo de personas que se requiere para llevar a cabo
un proyecto de la magnitud de un censo, requiere de un manejo
gerencial muy particular: autonomía, trabajo por resultados u
objetivos, altos niveles de coinpet 111 v 1 dad, etc. son condicio
nos propicias para atraerlas y mantenerlas.
Como conclusión, asi como se dedican esfuerzos para depurar
la metodología censal y la elaboración de preguntas del formu
lario, también se deberían mejorar en forma sistemática las
estrategias para garantizar el éxito de los censos. Se puede
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
Ill
tener la mejor metodología, los mejores demógrafos e ingenie-
ros de sistemas y la tecnología más avanzada, pero si no se
logra garantizar una recolección oportuna y completa, todo
ello será en vano.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
Una Mirada al Censo de 19B0 de los Estados Unidos de América.
Peter A. BOUNPANE
Bureau of the Census
Estados Unidos de América
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
El censo de la población y la vivienda de 1980 de los Estados Unidos fue tomado
el primero de abril de 1980, siendo asf el vigésimo censo en la historia de la
nación. En el censo de 1980 se registraron mas de 226 millones de personas y
más de 88 millones de unidades de vivencia. El documento adjunto (en ingles)
describe en detalle el censo de 1980. Éste es un resumen de la presentación
verbal que acompaña ra'el documento escrito. La presentación describe las
evaluaciones que se han llevado a cabo comenzando con el censo del 1950. En
aquel momento se determino que los empadronadores introducen errores sustanciales
en el censo. Se llevaron a cabo varios experimentos para determinar si la
información obtenida mediante autoenumeración es más exacta. Basados en los
resultados positivos de estos experimentos se comenzaron a llevar acabo pruebas
de censos por correo. Basado en la evaluación de estas pruebas, el censo de
1970 utilizo7 el correo para contar el 60% de la población. En 1980 el 95% de
la población fue empadronada por correo.
El levantamiento de un censo por correo requiere un listado para el despacho
postal que sea exacto y este'' al dfa. En 1980 este listado se construyó de
dos maneras. En las áreas urbanas se empezó con listados comprados a compañías
privadas que se especializan en listas de direcciones. Estas listas se
pusieron al día varias veces tanto por el servicio postal como por los empleados
del censo. En areas menos urbanizadas fue necesario contratar empadronadores
para crear las listas ya que las compañías privadas no tenían listas para estas
áreas. Estas listas también se pusieron al día. Se realizó un gran esfuerzo
para preparar listas exactas. De hecho casi el 10% del total del presupuesto
censal se invirtió en la preparación de la lista de direcciones.
Los cuestionarios censales se enviaron el 28 de marzo de 1980, cuatro días
antes del di a del censo. Se le pidióla las familias que llenaran los cues-
tionarios en sus casas (usando autoenumeracidn) y que las devolvieran por
correo. Dentro de l*»r dos semanas siguientes el 83% de "i ÜS hogares devolvieron
los cuestionarios por correo. Oespues se contrataron empadronadores (300,000)
para visitar los hogares que no contestaron y completar la entrevista. Aun
con el censo por correo los Estados Unidos necesitaron un gran numero de
empadronadores. Sin el censo por correo dudamos que se hubieran podido
contratar suficientes empadronadores para completar el censo a tiempo.
Todos los cuestionarios (devueltos por correo o llenados por 1 os empadronadores)
fueron revisados para determinar si contenían toda la información necesaria.
Se volvio' a establecer contacto (por teléfono o visita personal) con aquellos
hogares cuyos cuestionarios tenían un gran numero de preguntas en blanco para
obtener la información que faltaba.
Se contrataron empadronadores especiales a los cuales se les dieron
instrucciones diferentes para hacerce cargo de situaciones especiales de
vivienda, por ejemplo, bases militares, dormitorios de universidades,
prisiones, etc.
El 5% de la población se.empadronó usando el método tradicional de puerta a
puerta. Este es el método empleado por la mayoría de los países representados
en esta conferencia.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
114
Doce oficinas regionales permanentes mantuvieron el control del censo.
Para efectuar el empadronamiento, se establecieron 409 oficinas de distrito
temporales. El personal de estas oficinas estaba compuesto, casi en su
totalidad, por empleados temporeros.
Se utilizaron dos cuestionarios, uno corto y otro largo (cuestionario de
la muestra). El cuestionario corto contenía 7 preguntas para cada persona
y 10 relativas a la unidad de vivienda. Cinco de cada seis hogares recibieron
el cuestionario corto, excepto en ciudades bien pequeras (menos de 2,500 habi-
tantes), en las cuales uno de cada dos hogares recibió' el cuestionario corto.
El resto de los hogares recibió' el cuestionario de la muestra, que contenía
las mismas preguntas del cuestionario corto mas 20 preguntas adicionales por
persona y 16 preguntas adicionales sobre la unidad de vivienda.
Los cuestionarios se enviaron a una de tres oficinas de procesamiento al
terminarse todas las actividades de recopilación de datos. En estas oficinas
los datos fueron leídos por una computadora usando el sistema F0SDIC del
Negociado del Censo. F0SDIC es un sistema detector de marcas que requiere
filmar los cuestionarios primero. Luego la maquina F0SDIC lee directamente el
filme en la computadora.
El procesamiento se llevo a cabo en dos fases. Primero, todos los cuestio-
narios se filmaron para obtener la información de la forma corta. (Para los
cuestionarios de la muestra solamente se filmaron las dos primeras páginas.)
Luego, las anotaciones escritas en los cuestionarios de la muestra fueron
codificadas. En la segunda fase, todos los cuestionarios muéstrales se
volvieron a filmar completamente. El procesamiento requirió 6,300 empleados,
costcf $115 millones y duro 12 meses. El sistema de procesamiento de 1980
funcionó muy bien. Fue rápido y exacto. Mas aun, no se necesitó una gran
cantidad de empleados para entrar la información en la computadora directamente.
Hubo sin embargo algunos problemas. El uso de F0SDIC impone restricciones en
cuanto al diseno del uestionario y requiere un papel de alta calidad.
Además, el cuestionario necesita cierta preparación antes de ser filmado.
Para poder corregir errores, se tuvo que filmar de nuevo unidades de trabajo
completas, no solamente los jestionarios incorrectos. Finalmente, hubo
inconsistencias en la calidad de la codificación entre las tres oficinas de
procesami ento.
La evaluación y la experimentación son parte integrante del censo en los
Estados Unidos. La cobertura (o sea, cuántas personas se incluyeron en
el censo) es una evaluación clave. La mayoría de los países evalúan este
aspecto de su censo. La cobertura del censo de 1980 se evalué de dos
maneras. Primero, se hizo una estimación independiente del total de la
población utilizando datos pasados y registros administrativos (nacimientos,
defunciones, inmigración, etc.). Este método se llama análisis demográfico.
Comparando la estimación independiente con los resultados censales se obtiene
una medida de subcobertura. El método sin embargo, es imperfecto debido a
inconsistencias en las definiciones y de cobertura entre el censo y los
sistemas de registros.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
115
Segundo, se tomo una encuesta especial después del censo (llamada Encuesta
de Evaluación Post-Censal o EEP). Otra estimación de la cobertura se obtiene
pareando las personas incluidas en esta encuesta con las incluidas en el
censo. Este método es quizás el más común seguido por otros países. Esta',
por supuesto, sujeto a errores en el pareo.
Para el censo de 1980, mediante el análisis demográfico se estimo una subco-
bertura de 1.4%. Las estimaciones del EEP variaron entre 1% de sobreenumeracion
y un 2% de subenumeraci on, dependiendo de lo que se asumió'' con respecto a los
casos que no parearon. Estos niveles estimados de subenumeracion son muy
bajos; sin embargo, las estimaciones para subgrupos de la población varían
considerablemente con respecto a las estimaciones de la subcobertura global.
Ademas de la evaluación de la cobertura se hi,cieron también estudios para
medir la calidad de los datos recopilados. Éstos se hicieron principalememte
mediante la comparación de las tabulaciones censales con la información
contenida en registros administrativos independientes o volviendo a entrevistar
a una muestra en gran escala de hogares. En la reentrevista se hicieron las
preguntas censales nuevamente y se reconciliaron las discrepancias.
En los censo recientes se condujeron experimentos como parte del censo. Es
imposible reproducir las condiciones censales en los anos intercensales, de
manera que tiene sentido experimentar conjuntamente^con el censo. En 1980
los experimentos como parte del censo examinaron métodos nuevos para la
distribución de los cuestionarios, formatos alternativos del cuestionario y
un mayor uso del telefono en el proceso de la entrevista.
Las evaluaciones y los experimentos son parte esencial de un censo. Se
discuten las siguientes pautas para la evaluación:
- Deben ser planeadas con anticipación.
- Debe dárseles una alta prioridad.
- Se debe solicitar el consejo de otros campos profesionales relacionados
al planear las evaluaciones.
- Las observaciones mientras se esta llevando a cabo el censo son tan
importantes como las evaluaciones formales.
- Las evaluaciones deben ser súplementadas con pruebas intercensales.
- Los experimentos se deben conducir conjuntamente con el censo, siempre
que sea posible.
- Se debe aprender de las evaluaciones y los experimentos de otros países.
En conclusion, cada censo en cierta forma mira hacia censos del futuro.
Debemos hacer uso de la oportunidad para evaluar cada censo, no solamente
para determinar cuan bien se hizo, sino también para ayudarnos a planificar
censos futuros.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
116
UNITED STATES
OF AMERICA 1980
Introduction
The 1980 Census of Population and Housing— 20th in a
chain of censuses that have been taken every year ending
in zero since 1790—uas conducted as of 1 April 1980 by
the Bureau of the Census, an agency of the U.S. Depart-
ment of Commerce. The enumeration covered the popu-
lation and housing characteristics of the 50 states, the
District of Columbia, Puerto Rico, and the outlying
areas under U.S. jurisdiction or sovereignty (including
Guam, American Samoa, and, by special agreement,
the Northern Marianas and the Trust Territory of the
Pacific Islands). The 1980 census counted 226,545,805
persons and 88,411,263 housing units in the United States and 3,565,376 persons and 1,082,288 housing
units in Puerto Rico and the outlying areas.
History
The eminent nineteenth century French statistician Alex-
andre Moreau de Jonnés observed, "The United States
presents in its history a phenomenon which has no paral-
lel. It is that of a people who instituted the statistics of
their country on the very day when they formed their gov-
ernment."
De Jonnés's observation acknowledges the fact that
the U.S. Constitution, ratified in 1789, required an enu-
meration of the population at least every 10 years. The
constitutional mandate for a census stems from a com-
promise between the large and small states at the Con-
stitutional Convention in 1787. This compromise gave
each stale equal representation in the Senate, but linked
representation in the House of Representatives to each
state's population. Thus, article 1, section 2 of the Con-
stitution called for a census as the means of equitably
apportioning representatives among the states. This
constitutional mandate remains the primary reason for
conducting the census.
The first census was conducted in 1790 by U.S. mar-
shals and iheir assistants under the direction of the sec-
retary of state. The act authorizing this census, with
minor modifications and extensions, governed the tak-
ing of the censuses through 1840. The inquiries in 1790
were limited to six items—the name of the head of the family and the number of persons in each household of
the following descriptions: free white males 16 years and
older; free white males under 16; free white females; all
other free persons (i.e., free blacks); and slaves. By 1840
the marshals were also gathering data on education, lit-
eracy, and occupation, as well as taking censuses of agri-
culture, manufactures, and mineral industries. In 1850
the new census law provided for collecting the names
and characteristics of each person counted.
In 1880 specially appointed supervisors and enumera-
tors began taking the censuses instead of the U.S. mar-
shals and their assistants. The 1880 population census
was an encyclopedic undertaking that made hundreds of
minor inquiries in addition to the basic population ques-
tions. Because of the vast scope of this census, however,
publication of the detailed results was not completed
until just before the 1890 enumeration. The 1890 cen-
sus, which was similar in scope to the 1880, introduced
the use of punchcards and electric tabulating machines
to accelerate data processing.
The censuses from 1790 to 1900 were conducted by
temporary staffs that disbanded after the count was
completed and the results published. Some observers,
particularly in the latter half of the nineteenth century,
noted that this ad hoc census taking was inadequate in
several respects: there was a lack of continuity and expe-
rience in census work, the enumerations had to be
organized in haste, and the accuracy of the statistics w as
impaired. Based on these observations and recommen-
dations from government and the private sector, Con-
gress established a permanent Bureau of the Census
in 1902. With a permanent organization, the Census
Bureau was able to conduct lesser surveys and the cen-
sus of manufactures separately from the decennial cen-
sus.
The 1940 census was, in many ways, the first modern
census. One of the major innovations was the use of
sampling, which involved asking some of the questions
of only a fraction of the population. The 1940 census
From Censuses of Ada and the Pacific: 1980 Round, edited by Lee-Jay Oho and Robert L. Heam, © 1984, bast-West Population Institute, East-West Center, Honolulu, Hawaii. Reprinted with permission.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
117
v.as also the first lo obtain facts on the conditions of the
nation's housing.
The 1950 census brought the use of one of the first
electronic computers, UNI VAC I, lo help tabulate a
small part of the data. NearK aJI of the daia processing
for the 1960 census was done by computer, and an elec-
tronic device for reading the census schedules—called a
Him optical sensing device for input to computer, or
FOSDIC—was used in this and subsequent censuses.
Prior to 1960, census enumerators used a conven-
tional door-to-door procedure, completing a census
questionnaire at each housing unit. But in 1960 mail car-
riers delivered an unaddressed questionnaire containing
the basic 100 percent questions to every housing unit
before the field enumeration. The householder was asked to complete the questionnaire and hold it until an
enumerator visited to pick it up. In areas containing
about 80 percent of the population, the enumerator
picked up the 100 percent questionnaire, and, at every
fourth housing unit, left another containing the sample
population and housing questions requesting that the
respondent fill it out and mail it to the census district
office. When these questionnaires were returned, the
responses were transcribed to the special FOSDIC
schedules. In rural areas, the sample information was
obtained during the enumerator's visit and recorded
directly on FOSDIC schedules. Self-enumeration had
been used on a very limited scale previously, but this was
the first time it was made a major part of the decennial
procedure.
The mails were used even more extensively in the 1970
census. Approximately 60 percent of the housing units
received addressed questionnaires by mail and were
asked to complete them and mail them back. Some of
the households received short forms containing only the
100 percent questions, while households predesignated
on the address lists received long forms containing the
100 percent and additional questions. All the question-
naires were designed to be read by FOSDIC. In the areas
where this procedure was used, enumerators contacted
only those housing units and households for which
questionnaires were not returned, were incomplete, or
contained inconsistent answers. For the remainder of
the housing units, most of which were located in rural
areas or small towns, mail carriers left a census form
containing the 100 percent questions at each residential
housing unit on their routes, as in 1960. An enumerator
then visited to collect the completed questionnaire and
ask additional questions for the sample units identified
in the listing books.
In the 1970 census, changes in subject content over
1960 were relatively minor. The only population data
collected on a 100 percent basis related to the same five
subjects (age, sex, race, marital status, and relationship
to household head). The sample questions were asked of
either a 15 percent or a 5 percent sample of housing
units, with a number asked for both. Fifteen housing
items were covered on a complete-count basis; others
were collected on a sample basis similar to that used for
the population inquiries.
The method of enumeration for the 1980 census was
not radically different from that for 1970, though there
were some significant developments: (1) the mail-out/
mail-back area was extended to over 95 percent of the
housing units, (2) new procedures for improving the
coverage of the population were introduced and 1970
procedures were enhanced, and (3) the publicity cam-
paign was the most extensive and creative ever under-
taken for a census. The questionnaires contained about
the same number of items as the previous census, but
there were some subject changes reflecting new data
needs. The sample design was different (see Sample
Design section below), but the general approach of ask-
ing only a few basic questions of everyone and the more
deiailed items of a sample of the population was the
same as in 1960 and 1970.
Legal Status
The U.S. Constitution does not describe how the census
should be conducted, which questions are to be asked,
or other vital aspects of census taking; instead, it
empowers the Congress to conduct the census in "such
Manner as they shall by Law direct." Congress passed
special acts for the taking of each census from 1790-
1920 that gave quite detailed instructions about how
to conduct the enumeration and which questions to ask.
In 1929, Congress passed the Permanent Census Act, under which the 1930 census was taken. The most nota-
ble feature of this act was the discretion it gave to the
secretary of commerce and the Bureau of the Census
with regard to conducting the census. Modifications to
the 1929 act, and additional legislation covering the cen-
sus of housing, governed the 1940 and 1950 censuses.
The basic law under which the 1980 census was taken
was title 13 of the United States Code, which was passed
in 1954 and has been amended several times since. Like
the 1929 act, title 13 does not specify which questions
are to be asked, nor does it specify the method of enu-
meration. It does, however, contain provisions relating
to the areas to be covered, and it requires that the Cen-
sus Bureau advise Congress of the general subject con-
tent three years before the census and of the specific
questions two years before. Furthermore, it requires the
Census Bureau to deliver the state population counts
and apportionment results to the president "within 9
months after the census date."
Title 13 requires individuals to answer the census.
Anyone 18 years of age or older who willfully neglects
or refuses to answer the census may be fined up to $100.
Anyone who gives false answers is subject to a fine of up
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
118
iu C '00. I.i 197'\ 1;i a.! ' ' . 'o 'he fines, there were pro-
visions for minor jail terms for refusing to answer or
answering falsely, but these were dropped for 1980.
The same law that makes answering the census man-
datory provides strict confidentiality for the informa-
tion gathered. It states,
Neither ihe Secretary, nor any other officer or employee of
the Department of Commerce or bureau or agency thereof,
may . . .
(1) use the information furnished under the provisions of this title for any purpose other than the statistical pur-
poses for which it is supplied; or
(2) make any publication whereby the data furnished by any particular establishment or individual under this
title can be identified; or
(3) permit any other than the sworn officers and employees of the Department or bureau or agency thereof to exam-
ine the individual reports.
Every employee of the Census Bureau must take an
oath to protect the confidentiality of information
gathered in the census. Any employee who wrongfully
discloses census information is subject to a fine of up to
$5,000 and imprisonment up to 5 years.
Census records are by law confidential for 72 years
from the time the information was collected. Many peo-
ple rely on copies of their census records to prove age or
identity, and the Census Bureau releases such informa-
tion only to those persons, their authorized representa-
tives, or their legal beneficiaries upon proof of death.
The Census Bureau was required by law to provide
the president, by 1 January 1981, with the final official
state population counts from the 1980 census and the
number of representatives to which each state was
entitled in the House of Representatives. Related to the
apportionment of representatives is the delineation of
congressional and legislative district boundaries on the
basis of "one person, one vote" (the concept that legis-
lative districts should have nearly equal populations).
Under the provisions of a law enacted in 1975 (P L. 94-
171), the bureau was required to produce population
data for geographic areas outlined in plans submitted by
the state offices or public bodies that had initial respon-
sibility for t he legislative apportionment or districting of
the state (the secretary of commerce established the cri-
teria for the state plans). These data had to be delivered
to the states by 1 April 1981. It should be noted that,
while the Census Bureau provided data that could be
used for redistricting, it did not actually draw the con-
gressional or legislative district boundaries.
The decennial census has, of course, many uses other
than apportioning seals in the House of Representatives
and drawing legislative district boundaries. Census data
are also used for allocating federal and state funds
under v.ihous prants in-aid and revenue-sharing pro-
grams, for formulating public policy, and in pri\atc-sec-
tor decision making. Legislation passed during the
1970s tied the djitiibuiion of federal rc\cnue sharing
funds to population totals and characteristics; this
action contributed to a heightened public awareness of
and support for the 1980 census.
Organization of the Bureau of the Census
The bureau's headquarters is in the Washington, D C.,
suburb of Suitland, Maryland. It has processing and
operational offices in Jeffersonville, Indiana, and Pitts-
burg, Kansas, where large-scale clerical operations are
conducted, and regional offices in 12 cities through-
out the country—Atlanta, Boston, Charlotte, Chicago,
Dallas, Denver, Detroit, Kansas City (Kansas), * Los
Angeles, New York, Philadelphia, and Seattle (see fig-
ure 19.1). To conduct the field enumeration, a regional
census center (RCC) was set up in each regional office
city. The 12 centers directed the work of 409 temporary
district offices in the 50 states and the District of
Columbia, 8 in Puerto Rico, and one for each of the
outlying areas.
Two large processing sites were set up for the 1980
census operations, one in New Orleans, Louisiana, and
the other in Laguna Niguei, California. These offices
closed in early 1982. The Jeffersonville facility was also
a major site of 1980 census processing.
The field enumeration, or data collection, was the
direct responsibility of the bureau's Field Division at
headquarters, and the RCCs were responsible for direct-
ing operations in the temporary census district offices.
RCC personnel trained key district office supervisors,
monitored costs and progress of operations in the dis-
trict offices, processed district office payrolls, and in
general had overall responsibility for assuring the timely
completion of fieldwork. The person responsible for
directing the operation in the RCC was the regional cen-
sus manager (officially called the assistant regional
director—census), who was assisted by technical special-
ists for operations, administration, geographic matters,
recruitment, publicity, and community services. The
regional census manager reported to the regional direc-
tor of the permanent regional office, who in turn re-
ported to the Field Division at headquarters.
There were four types of district offices. In mail-cen-
sus areas, district offices were either "centralized" or
"decentralized." The 87 centralized offices were in
inner-city areas, while the 286 decentralized offices were
located primarily in suburban and rural areas. There
were 24 offices in areas of the country where the con-
ventional method of enumeration was used. In addition,
there were 12 two-procedure offices where both conven-
tional and decentralized procedures were employed.
District office operations were under the direction of
UNITED STATES INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
Figure 19.1. 1980 L'.S. Regional Office Boundaries 119
NOTE: During the 1980 census, there were two regional boundary systems—one for activities that had been operational before the 1980 census (current surveys, informational services, etc.) and one for decennial activities only. The decennial boundary system was in use roughly from 1978 to 1981.
a district office manager, who was assisted by top-level
supervisors for field, office, special place, administra-
tive, and recruitment operations. The census enumera-
tors worked under the supervision of crew leaders and
higher-level field supervisors, and there wa«; a clerical
force in each office.
Planning, direction, and support services were pro-
vided by the bureau's permanent staff. This staff was
augmented for the peak periods of census operations—
particularly in the areas of training, publicity, and pro-
cessing—and then reduced as operations came to a
close. Large numbers of temporary workers were hired
for the field offices and processing centers.
Design and Execution
The 1980 census involved several overlapping phases,
extending generally over the period 1973-83 and involv-
ing planning and preparation, data collection, data pro-
cessing, data dissemination, and evaluation.
The 1980 Census of Population and Housing also
included two small surveys—the Components of Inven-
tory Change Survey, which was designed to obtain
information on counts and characteristics of housing
units that changed or stayed the same between 1973 and
1980, and the Residential Finance Survey, which was a
survey of residential properties to obtain data on mort-
gages, shelter costs, and selected housing and owner
characteristics.
Preparation and Planning
Planning for the 1980 census began while the last phases
of the 1970 census were still underway, and funding for
formal planning began in July 1973. The planning pro-
cess included a review of the experiences in the 1970 cen-
sus, consultation and contacts with data users, congres-
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
United States- Johnston Atoll, Midway, \Vake, and
miscellaneous otner islands--wee either unirfndñtcd or
had counts svpphed for them by other j'cdeial ;,g. k ;es.
All persons living in the United Slates on ccr-us day
were covered in the census, including foreigners having
their usual residence in the United States, whether
legally 01 illegally. (While illegal 'aliens wtie to be
counted in the census, no attempt was made to identify
them as such.) Included weie persons working or
attending school in the country and members of their
families living with them, Foreigners temporarily visit-
ing or traveling in the United States or living on the
premises of an embassy, ministry, legation, chancellery,
or consu!. te were not enumerated. Procedures were
established to count U.S. residents who v.ere short term
travelers abroad, but U.S. citizens residing o.erseas
(including federal civilian employees, students, and
armed forces personnel stationed there) were excluded.
As in every preceding census, each person was
counted as an inhabitant of his or h,r usual p!ai.c oí' res-
idence, which generally meant the place where the per-
son lived and slept most of the tin".:-. This was not neces-
sarily the same as the person's legal or voting residence.
Rules were established for enumerating categories of the
population for whom residence was not obvious.
The basic unit of enumeration was the housing unit.
Population characterises were obtained for each person
living in an occupied housing unit (a household) and
housing characteristics were obtained for both occupied
and vacant units. A housing unit v as defined as a
house, an apartment, j group of rooms, or a single
room occupied (or if vacant, intended for occupancy) as
a separate living quarters. Separate living quarters were
those in which the occupants or intended occupants
lived and ate separately from other persons in the build
ing and which had direct access from the outside of the
building or through a common hall Boats, tents, vans,
caves, and the like were included in the housing inven-
tory only if they were occupied as someone's usual place
of residence. Vacant mobile homes were included pro-
vided they w;ere intended for occupancy on the site
where they stood.
If a living quarters contained nine or more per: cms
unrelated to the resident owner or renter, or 10 or more
unrelated persons, it was considered a "group quar-
ters." College dormitories, military installations, pris-
ons, hospitals, orphanages, convents, and "the hke i'cil
under this category. Areas with group quarters were
called "special places"' (these sometimes contained regu
lar housing units as well). Characteristics were collected
for the group quarters population in a separate opera-
tion. No housing characteristics data were collected on
group quarters.
Members of the armed forces living on military instal-
lations were counted, as in every previous census, as res-
idents of the area in which the base was located, mem
120 bers not living on base were counted as residents of the
area m which they were li\ ing on census day. Persons in
families with aimed forces personnel were counted
where they were living on census day (i.e., on the mili-
tary installation or off ba^e, as the case might be).
Fach naval ship w as allotted to the municipality that
the Department oi the Navy designated as its home
pon, except for those of the Sixth and Seventh Fleets,
which were deployed overseas on census day (these were
considered part of the overseas population because of
their long-tetm assignments). In home ports with fewer
than 1,000 naval personnel assigned to ships, the crews
were counted aboard the ship. In home ports with 1,000
or more naval personnel assigned to ships, the personnel
who indicated that they had a usual residence within 50
miles of the home port of their ship were attributed to
that residence, and those who did not so indicate were
counted aboard ship. When a home port was split
between ■- icipalities, ships berthed in the home port
on . u, ..:, day were assigned by the bureau to the one in
which the ¡and immediately adjacent to the dock or pier
was located. Ships not physically present in their home
port and not deployed to the Sixth or Seventh Fleets on
census day were allocated to the municipality named on
the navy's home port list.
If a U.S. merchant vessel was berthed in a U.S. port
on census day, the crew was enumerated at that port. If
the ship was not docked in a U.S. port but was inside the
territorial waters of the United States, the crew was enu-
merated at the port of destination, if in the United
States, or at the home port of the ship if its destination
•- -r- outside the United States. Crews of U.S. flag vessels
that were outside American waters on census day and
crews of vessels flying a foreign flag were not enumera-
ted in the 1980 census.
College students were counted as residents of the
areas in which they were living while attending school,
as they have been since 1950; however, students in
boarding schools below the college level were counted at
their parental homes.
Inmates who ordinarily lived in an institution for con-
siderable periods of time were counted as residents of
the area where the institution was located. Patients
in general hospitals for a short period of time were
counted at their usual place of residence; if they had
none, they *ere counted at the hospital.
Persons in hotels and motels on the night of 31 March
¡980 were requested to fill out an individual census
report and were assigned to their home areas if they
indicated that no one was at home to report them to the
census A similar approach was used for persons visiting
in private residences, as well as for Americans who left
the United States during March 1980 via major inter-
continental air or ship carriers for temporary travel
abroad. In addition, information on persons away from
their usual places of residence was obtained from their
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
121
sional review, and a series of tesis of procedures and
III.
The exlensivc use of the mails in )970 had proved suc-
ce< rnl. About 86 percent of the households in mail cen-
sus areas cooperated by mailing back their question-
naires to the census district offices. The Census Bureau
conducted a test during the 1970 census of the feasibility
of extending the mall method to rural areas and small
towns. Based on the 1970 mail-return rates, the results
of the mail extension test, and other factors, a decision
was made to widen Ihe mail-census areas in 1980 to
include 95.5 percent of the population. Another aspect of the 1970 census that affected plan-
ning for 1980 was the estimate that about 5.3 million
people, or 2.5 percent of the population, had been mis-
sed. Further evaluations indicated that blacks had been disproportionately undercounted: while only 1.9 per-
cent of whites were not counted, it was estimated that
7.7 percent of blacks were missed. (A reestimation of
the 1970 census undercount done in 1981 with new fig-
ures for such factors as emigration estimated that 4.7
million people were missed in 1970, or 2.2 percent of the
population. The undercount for blacks was 7.6 percent
and for whites, 1.5 percent.) Limitations in the adminis-
trative data (birth, death, and migration records) used
to estimate population, and thus undercount, did not allow calculating the undercount for other minority
groups—Hispanics, American Indians, and Asians and
Pacific islanders.
No issue figured as prominently in planning the 1980
census as the undercount of minority groups in 1970.
Considerable time and money went into identifying
methods for improving coverage of these hard-to-enu-
merate segments of the population. The goals for cover-
age improvement in the 1980 census were (1) to attain a
relatively low overall undercount and (2) to reduce the
coverage differential between whites and minorities.
The Census Bureau took two main approaches
toward achieving these goals. First, it undertook a num-
ber of special publicity and outreach efforts to make
people more aware of the census, to explain the impor-
tance of census data, and to convince people that their
responses were confidential. Special efforts were
designed to reach minority groups. One of these was the
Community Services Program, in which over 200 com-
munity services specialists contacted leaders of commu- nity-based minority organizations and American Indian
tribes to obtain their active support for the census.
The second approach was to improve census-tak-
ing procedures to reduce the possibility of people being
missed. The major improvements were (I) additional
checks on the compilation of address lists to be used as
controls in the mailing and handling of the census ques-
tionnaires, (2) matching to census records the names of individuals living in selected hard-to-enumerate areas
(the names were taken from drivers' license lists), (3)
rechcckinc the occupancy status of units that had been
classified vacant or nonexistent, (A) establishing special
procedures for counting the transient populations, and
(5) ¿;i\ing ..U ¡ ¡icials .■ .ppoitunit) to rc\iew census
counts before the census district offices were closed.
The Census Bureau consulted a broad spectrum of
data users in planning the 1980 census, and it received
advice from a dozen public or other advisory commit-
tees. These included members of the American Statisti-
cal, American Economic, and American Marketing
Associations, committees on population and housing,
and three minority advisory groups representing blacks,
Hispanics, and Asian and Pacific island Americans.
Regional meetings held with American Indian and
native Alaskan groups were forums for an exchange of
ideas on how best to count these people. Representatives
of more than 90 federal agencies formed the Federal
Agency Council for Demographic Censuses, which out-
lined federal data needs, helped advise on census con-
tent, and reviewed other matters related to the census.
Participants in local public meetings held throughout
the country suggested improvements for the 1980 count.
Meetings were also held with state planning agency offi-
cials to get their views. A blue-ribbon panel of the
National Academy of Sciences/National Research
Council reviewed census plans and made recommenda-
tions. Extensive contacts were made with national and
community-based minority organizations to inform
these groups of plans and to gather comments.
In addition, the Census Bureau's plans for the 1980
census underwent extensive congressional review.
Numerous hearings were held before the bureau's
House oversight committee (the Subcommittee on Cen-
sus and Population of the Committee on Post Office
and Civil Service) and Senate oversight committee (the
Subcommittee on Energy, Nuclear Proliferation, and
Federal Services of the Governmental Affairs Commit-
tee). The General Accounting Office conducted and
published a number of studies of 1980 census plans.
Plans were also reviewed by the House and Senate com-
mittees responsible for funding the census.
Coverage
The territory covered by the 1980 census included the 50
states and the District of Columbia (these 51 units com-
prise the official population of the United States),
Puerto Rico, and the outlying areas—Guam, the U.S.
Virgin Islands, American Samoa, the Commonwealth
of the Northern Mariana Islands, and the Trust Terri-
tory of the Pacific Islands. As a result of the 1978 treaty
between the United States and Panama, the Canal
Zone, which had been enumerated in each census from
1920 to 1970, was not part of the 1980 census. A number
of other areas under the jurisdiction or control of the
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
122
were not defined. In all parts of the country they were
the administrative units to be covered by census enu-
merators, generally one ED per enumerator.
Maps are essential tools for both data collection and
data dissemination. For the 1980 census, the bureau
produced more than 32,000 individual maps covering
the entire country, Puerto Rico, and the outlying areas.
The bureau started with existing maps from the U.S.
Geological Survey, state highway departments, county
governments, and incorporated places and adapted
these for the census by removing unnecessary informa-
tion and adding boundaries for special statistical areas.
Enumerators were given maps with their assigned
EDs delineated in order to keep them within their areas
and to assure that they found every housing unit and
assigned each unit and its inhabitants to the correct geo-
graphic area. *
The maps used for fieldwork reflected the geographic
situation as of 1 January 1978 for mail-census areas and
1 January 1979 in door-to-door areas. During the field
enumeration, the maps were updated to reflect current
conditions. The official date for census geography was 1
January 1980; any changes effective after that date,
such as annexations, were not reflected in the final cen-
sus geography.
Updated maps were made available to users so they
could relate the data to the proper geographic area.
Appropriate maps were also included with the printed
and microfiche reports.
Schedule Content
Census content, or what questions would be asked, was
a major focus of 1980 census planning. The bureau's
goals were to meet the data needs of the 1980s without
unnecessarily burdening the public. As in 1970, most
questions were asked of only a sample of the popula-
tion, and only a few basic population and housing items
were asked of everyone. The 1970s ushered in new con-
cerns and data needs (e.g., more detailed data on minor-
ity groups, transportation, and housing costs); thus,
some of the questions asked in 1970 were deemed no
longer necessary and new ones were requested by vari-
ous sources. Some questions were added, while some
1970 items were dropped. Similarly, some questions
asked on a sample basis in 1970 were asked on a 100 per-
cent basis in 1980, and vice versa.
Table 19.1 lists the 1980 census content items (see
appendix 1 for a facsimile of the long-form question-
naire). The term "head of household" was not used in
the 1980 census. Instead, the census questionnaire
instructed respondents to enter in column 1 the name of
one of the household members in whose name the hous-
ing unit was owned or rented, and then asked for each
persons' relationship to the person in column 1. Exam-
ples of items that can be derived from the census ques-
tions are families, family type and size, family income,
poverty status, population density, household size, per-
sons per room (overcrowding), institutions and other
group quarters, gross rent, and farm residence.
There was space on both the short and long forms for
enumerating up to seven people in each household. If
there were more than seven people in a household, the
respondent was to list all the persons in question I, com-
plete the form for seven people, and mail it in. Enu-
merators obtained information for the additional per-
sons in a follow-up visit.
Questionnaires for individuals ("individual census
reports"), containing only population inquiries, were
used for persons in group quarters. The information on
the individual census reports was later transcribed onto
regular FOSDIC-readable questionnaires. A supple-
mentary questionnaire was used at some American
Indian reservation households and at some households
in areas of Oklahoma that were formerly Indian reser-
vations to get more detailed information about special
living conditions of native Americans. Spanish-lan-
guage versions of the main short- and long-form ques-
tionnaires (as well as questionnaires for individuals)
were available upon request. In addition, translations of
the short- and long-form questionnaires were prepared
in 32 different languages. Separate questionnaires were
developed for Puerto Rico and each of the outlying
areas (with common subject content for the Pacific
jurisdictions). No sampling was done in the outlying
areas because their small populations were not condu-
cive to sampling methods. Instead each housing unit
received a questionnaire similar in length to the stan-
dard long form. Most types of questionnaires were spe-
cially designed so that when they were microfilmed, the
film could be read by FOSDIC.
Pretests and Pilot Censuses
One of the most important components of the planning
for the 1980 census was the series of pretests and dress
rehearsals conducted between 1976 and 1979. The pre-
tests were designed to examine the feasibility and cost-
effectiveness of new and alternative field operations,
enumeration procedures (particularly those designed to
improve the coverage of the population), and question-
naire content items. The major tests were the National
Content Test (1976), which was devoted entirely to test-
ing alternative question formats and wordings, and tests
primarily of census procedures in Travis County, Texas
(1976), Camden, New Jersey (1976-77), and Oakland,
California (1977). The latter three pretests were "mini-
censuses" in which most facets of enumeration were
studied. In addition, there were a number of other tests,
some held as early as 1975, which were designed to try
out specific procedures or questionnaire content items.
The dress rehearsals (pilot censuses) were the final
run-throughs of procedures planned for the 1980 cen-
sus. These were conducted in 1978 in the Richmond area
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
families, resident managers, or neighbors. If an entire
household was expected to be away during the whole
period of the enumeration, information on that house-
hold was obtained from neighbors. A matching process
was used to eliminate duplicate reports for persons who
reported for themselves while away and who were also
reported by someone else.
A special enumeration was conducted on the night of
6 April 1980 in such places as missions, flophouses, jails, and detention centers, and persons enumerated
there were counted as residents of the area in which the
establishment was located.
Sample Design
Ttoo main questionnaires were used in the 1980 census—
the short form containing the basic, or 100 percent,
population and housing questions asked for all persons
and housing units, and the long form containing the
basic items plus questions only asked of a sample of the
population and housing units. One housing unit in six
was required to answer the long form, except for gov-
ernmental units estimated to have under 2,500 people,
where one half the housing units were sampled. This
created a 16V, percent sample in larger areas and a 50 per-
cent sample in smaller areas, and meant that about one-
fifth of the population nationwide was enumerated on the
long form. The 50 percent sample for governmental units
with under 2,500 people made it possible to produce sub-
stantially more reliable data from the sample questions
than would have been possible with the 16Vj percent sam-
ple used elsewhere.
Geographic Structure and Mapping
The 1980 census provided data for numerous political
and statistical geographic areas. The political areas
include the states, counties, county subdivisions, and
incorporated places. Areas that have been specially
created for statistical purposes include standard metro-
politan statistical areas, urbanized areas, census county
divisions, census tracts, and city blocks. Below are definitions for some of the areas for which
census data are reported. Data for the 100 percent ques-
tions are published for all census areas down to the
smallest blocks, while data for the sample questions are
published only at the census-tract or block-group level
and above.
United States. The 50 states and the District of
Columbia.
Regions and Divisions. There are four census regions
(West, South, Northeast, and North Central) defined
for the United States, each composed of two or more
geographic divisions. The nine divisions are groupings
of states.
Political Units. The states, District of Columbia,
Puerto Rico, other outlying areas, congressional dis-
tricts, counties or county equivalents, minor civil divi-
123
sions (MCDs) such as towns and townships, incorpo-
rated places (e.g., cities or villages), and, in some states,
election precincts. In addition, data are provided for
American Indian reservations and Alaska native vil-
lages.
Standard Metropolitan Statistical Areas. SMS As for
1980 comprised one or more counties defined around a
central city of 50,000 population or an urbanized area
of 50,000 or more inhabitant's with a total metropolitan
population of at least 100,000. Contiguous counties are
included if they have a high degree of social and eco-
nomic integration with the area's population nucleus. (New England SMSAs are defined in terms of towns and
cities, rather than counties.)
Standard Consolidated Statistical Areas. SCSAs were
composed of two or more closely related SMSAs having
a combined population of 1 million or more.
Urbanized Areas. An urbanized area consists of a
central city and surrounding densely settled territory
with a combined population of 50,000 or more inhabi-
tants.
Urban/Rural. The urban population comprises all
persons living in urbanized areas and in places of 2,500
or more inhabitants outside these areas. All other popu-
lation is considered rural.
Metropolitan/Nonmetropolitan. Metropolitan areas
are inside SMSAs; nonmetropolitan areas are outside
SMSAs. Both metropolitan and nonmetropolitan areas
can contain urban and rural population.
Census County Divisions. CCDs were created by Cen-
sus Bureau and local officials in 20 states in which the
MCDs were not adequate for reporting census statistics.
Census Designated Places. CDPs, formerly referred
to as "unincorporated places," are closely settled popu-
lation centers without legally established limits. They
were delineated with state and local assistance for statis-
tical purposes, and generally have a population of at
least 1,000.
Census Tracts. Statistical subdivisions of counties.
Each tract, averaging 4,000 inhabitants, is delineated by
local committees (subject to Census Bureau standards)
within SMSAs and other selected areas.
Blocks. Generally bounded by streets or other physi-
cal features, blocks are defined and numbered in
urbanized areas, incorporated places of 10,000 or more
population, and other areas that contracted with the
Census Bureau for the collection of block statistics.
Block Numbering Areas. BNAs were defined for the
purpose of grouping and numbering blocks where tracts
were not established.
Block Groups. BGs were sets of contiguous blocks
within a census tract or BNA and were used in lieu of
enumeration districts for tabulation purposes in blocked
areas. Enumeration Districts. EDs were used for census tab-
ulation purposes where census blocks and block groups
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
f romo:;-■■na! n c-■ -v * :it.es ;:k!uc!.:;c :hc i üb'ic lnfor-
ir, it)> i Office .iJ :hí •••cIJ [\»¡<ion. Ü^nad planning
¡ or promotion o»" the ]•" tensu> tejar early i?, .he dec-
ade, ,.nd a of pioj^is conJc-U-d bctwe.-n 1972
and 1978 aidea this, including the pretests and pilot
censuses.
The CPO secured the free services of the Advertising
Council in directing a major media advertising cam-
paign. The decision to use free "public service" adver-
tising rather than to seek funds from Congress for a
paid campaign was controversial: some observers
doubted the effectiveness of a free effort. But the
Advertising Council's campaign, developed by the firm
of Ogilvy & Mather, proved to be a great success. An
independent study found that the commercial dollar
value of the advertising received by the bureau in the
period between January and June 1980 was nearly $38
million. The bureau's own publicity budget covered
such services as photography, filming, and graphic
artwork, and a service charge for Advertising Council
office and processing operations.
The campaign was conducted in ail major media: tele-
vision, radio, newspaper, transit cards, outdoor bill-
boards, business and trade press, etc. The centerpiece of
the promotion effort war, the slogan "Answer the Cen-
sus, We're Counting on You."
A committee of leaders in the broadcasting industry
representing all major markets in the country was
formed to ensure that the census messages were aired on
the radio and television stations in their areas. A num-
ber of celebrities, including some well known among
minority communities, donated their time to tape public
service announcements (some in Spanish and Chinese)
or to undertake other efforts endorsing the census and
urging audience cooperation.
The advertising campaign was only part of the
bureau's 1980 promotional effort. Census information
kits were mailed to more than 44,000 magazines and
to 22,000 newspapers and television and radio sta-
tions. The latter were tailored specifically for the appro-
priate state and type of media. Special kits were
designed for black, Hispanic, American Indian, and
Asian and Pacific islander news media. Another kit was
given to each member of Congress.
The director of the Census Bureau sent a letter to the
chief executive officers of the nation's largest corpora
tions requesting their assistance in supporting the census
promotion. As an example of the response, the Ameri-
can Telephone & Telegraph Company included census
mess?.ees with 90 million telephone bills sent out in Marcñ 198(.-. The director also wrote to .*,00 national
organizations soliciting assistance through their mem-
be: "hips; the Boy Scouts, for instance, del'vered 30 mil-
lion census '.-lochores dooi-to-dooi in M.'c't' >"80. As
; n e.-:?jnple of i . ovcninentnj coopciition, March
124
19S0 -,oc:al security check mailings also contained cen- sus promotion^ menaces.
Ki:s containing reprcJjciKe rv.teriaL were ser'
to over 100,00) elementary .vnd secondary school
throughout the country. The goals of i his program were
to have younger children involve their parents and other
family members in the census through take-home
assignments and to prepare teenagers in households
with language and reading difficulties to assist in filling
out the census questionnaires.
During the field enumeration, a 40-person network of
public-relations specialists operating out of the regional offices and district offices in major metropolitan areas
handled a wide variety of promotional responsibilities:
they obtained time for public service announcements on
local radio and television stations, advised the census
district managers on how to work with the press,
achieved the cooperation of local companies in the pro-
motion effort, and served as liaisons with complete-
count committees. At the urging of the Census Bureau,
more than 4,000 complete-count committees were
organized by local jurisdictions throughout the country
in an effort to generate local publicity. Census district
office managers and community services specialists also
played a role in disseminating the census message.
Enumeration
Census day was 1 April as it had been in each census
since 1930. Prior to that several dates were tried—
the first Monday in August was used from 1790 to 1820,
1 June from 1830 to 1900, 15 April in 1910, and 1 Janu-
ary in 1920. The first of April has been deemed the most
suitable reference date because of weather conditions
and the likelihood that people will be at their usual place
of residence—winter weather would impede the enu-
meration in some areas if census day were held much
earlier, and people are more likely to move or be away
on vacation in the summer months.
In order to complete the field enumeration before the
spring thaw in nothern and western Alaska, census day
there was 22 January 1980. As part of the agreement
with the local government, census day for the Trust Ter-
ritory of the Pacific Islands was 15 September 1980; this
date was chosen in order to facilitate the use of teachers
as enumerators. The Northern Mariana Islands were
pan of the Trust Territory of the Pacific Islands at the
time of the census, but were treated separately for pur-
poses of collection, tabulation, and presentation of cen-
sus data. Their census day was 1 April 1980. Questionnaires were generally to be completed gi- ing
mation as of 1 ApJi. even if they 'vne fi!v r¿t
days or weeks later. Several questions rc/cired to a per-
son's activity or condition at a point in time other than 1
April '980 (e.R., "Wkc :his persor ]:ve five y -rs
ago (Apr:! i, 1975'? ' ;iLh hr, icrscr ..ork at any
UNH ~D 1 TrvTES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
125
Table 19.1 1980 U.S. Census Content
Population Items
•Household relationship Sex
•Race Age Marital status
•Spanish/Hispanic origin or descent
Housing Items
i 00 percent
Number of housing units at address •Complete plumbing facilities Number of rooms in unit Tenure (whether the unit is owned or rented)
•Condominium identification •Value of home (for one-family owner-occupied units
and condominimus) Rent (for renter-occupied units) Vacancy status Duration of vacancy
School enrollment Educational attainment State or foreign country of birth Citizenship and year of immigration
••Current language and English proficiency ••Ancestry
Place of residence 5 years ago Activity 5 years ago Veteran status and period of service
•Presence of disability or handicap Children ever born Marital history Employment status last week Hours worked last week Place of work
••Travel time to work •Means of transportation to work
••Persons in car pool Year last worked Industry Occupation Class of worker
•Work in 1979 and weeks looking for work in 1979 •Amount of income by source in 1979
•Changed relative to 1970 "New item for Í980
Sample
Number of units in structure Stories in building and presence of elevator Source of water Sewage disposal Year building built
•Year moved into this house Heating equipment Fuels used for home heating, water heating, and cooking
•Costs of utilities and fuels •Complete kitchen facilities Number of bedrooms Number of bathrooms Telephone Air conditioning
•Number of automobiles ••Homeowner shelter costs for real estate taxes, fire and
hazard insurance, and mortgage
of Virginia, in two counties in southwestern Colorado,
and in a section of lower Manhattan in New York City—
areas representing the different sorts of conditions to be
encountered nationally. Efforts were made to keep the
testing of new procedural and questionnaire content
alternatives to a minimum with the intention of chang-
ing only those methodologies or questions that proved
problematic. Some procedural and questionnaire con-
tent changes did have to be made as a result of the dress
rehearsal experiences.
There was an extensive evaluation process for the
tests, consisting of formal statistical calculations, time
studies, reports based on staff visits to field offices, and
reviews at headquarters.
Publicity
The 1980 census promotional campaign was the most
creative and effective publicity effort in U.S. census his- tory. Its main focus was to inform the public, whose
cooperation is essential to the success of any census, of
the importance of census data and of achieving as com-
plete a count as possible. It was aimed at encouraging
those living in the United States to fill out their census
forms and, in mail-census areas, to mail them back to
the census district offices.
The promotional campaign was directed by the Cen-
sus Promotion Office (CPO), which was established in
the summer of 1978. Other bureau units also had census
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
]o^cr-Ie\e! yjpcr.rors (such a\ crew leaders) and non-
supervisor) e.nplo> ees. The ¡cms we e designtd to d(.;cr-
rfiine whether candidates could perform census-related
tasks. A Spanish-language version of the I<mer-!c\cl test
was made available in some areas.
In all, 1.2 million persons were tested for census jobs,
and about 80 percent of these persons passed. Approxi-
mately 460,000 persons were employed at one time or
another in the census district offices, with 270,000
working at the peak of activities in April and May 1980.
These positions included enumerators, crew leaders,
office clerks, and supervisory personnel. The RCCs had
about 1,800 people at the peak of activities.
The verbatim training method was used in 1980 as in
previous censuses; however, greater use was made of
audio-visual materials and learn by-doing exercises.
Training guides, which were to be read aloud word-for-
word by the trainer, were prepared in the bureau's Field
Division for each census position. Each trainee received
various aids and workbooks as supplements to the ver-
batim instruction. For the most part, each employee was
trained by the person who would be his or her supervi-
sor; thus, enumerators were trained by their crew
leaders, crew leaders by their supervisors, and so on.
After training, employees were to consult procedural
manuals for guidance on how to complete specific tasks.
Temporary employees were paid weekly in centralized
offices and biweekly in decentralized and conventional
offices. The pay rates varied by type of office, with the
highest rates in centralized offices and the lowest, in
conventional. Enumerators were generally paid on a
piece-rate basis, that is, a certain amount per short form
completed and a bit more per long form. The piece rates
were set sa that enumerators could earn a targeted
hourly wage of $4.00-$4.45. Under certain circum-
stances, enumerators were paid an hourly wage and
received mileage or iime-enroute payments. Crew
leaders were paid $4.50-55.10 hourly. Pay for of lice
clerks ranged from $3.55 to $3.75.
Enumeration in Mail-Census Areas
Offices in mail-census areas officially opened 2 January
19S0, though most were open for receipt of the first
truckload of supplies in mid or late December 1979.
Bureau staff leased nearly 4 million square feet of space.
Space for training crews of enumerators, which nor-
mally could not be handled in the district office, was
obtained rem free in schools, churches, post offices,
and other community meeting areas. Much of the activ- ity :n the dir.rin offi es prior to cenrus day v/as centered
around l;'.y."ne out office space and sr.ting u-j furniture,
organizing and inventorying supplies, and hiring staff.
A masí.ive logistics effort - involving the development
c; specifications, and the purchasing or leading and dis-
tribution of needed iríais—was re/wired to .uock each
126
off-co w¡;t| ;hc equipment and supplies needed to con-
duct !hc census. Each office was sent 1'4 to 2 tons of
supplies. Private co.-.tractors printed more than 170 mil-
lion short- and lon^-form questionnaires for use in the
census mail-out, enumerators' kits, and training. Other
contractors stuffed about half of these questionnaires
into specially designed envelopes along with the instruc-
tion booklet and return envelope. Some 75 million of
these packages were machine-labeled for each address
known to the bureau prior to the census; additional
questionnaires were addressed by hand in the district
offices following various mailing-list improvement
operations.
Items required in each office included the census
questionnaires, address registers, kits containing train- ing materials, procedural manuals, enumerator sup-
plies, and a number of special operational forms. Over
2,500 special forms were designed and printed for the
1980 census, and more than 2 million specially designed
kits were assembled at the bureau's processing offices
and shipped to the district offices. In addition, each
office received typical office supplies and equipment
such as filmstrip projectors and cassette players for use
in training. All offices were furnished with folding
chairs and cardboard tables, desks, and filing bins, and
a small amount of more durable furniture for the super-
visory staff. Typewriters were rented locally. Telephones
were installed as needed, including extra lines to handle
calls from respondents needing help in completing cen-
sus questionnaires and, in inner-city offices, lines for
calling respondents who returned incomplete question-
naires. Facsimile machines for transmitting population
counts and other information to the regional offices
were installed in most district offices.
One of the most important early jobs in the mail-cen-
sus district offices was improving mailing lists. This was
done to ensure that the mailing of questionnaires would
be as complete as possible. Each district office received
address registers listing addresses for each ED in its
area; these had been compiled in late 1979 and early
1980. In city-delivery areas, mailing lists used to compile
the registers had been purchased from private compa-
nies and had undergone an advance check by the U.S.
Postal Service in June of ¡979. The lists were checked
twice more by the Postal Service after the district office
opened, in early March 19S0 and at the time of delivery
of the questionnaires (28 March). In addition, census
enumeiators had conducted £ further check beginning
in mid-February in an operation called the precan-
vasi. Nccrssaiy additions deletions, cr corrections of
addrec .fs generated by 1 .u *ai rhecks and the precan- vass were made by clerks in the census offices. For areas
where mailing Iiif could r zl b: purchased from private
compr ,ie£, they were cou piled by census enumerar "
in th. spring ard summer of 197? in a "ureiist" cr v
I<\ iXSTvTEb
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
last week?" or 'Touring 1979 did this person receive any income from the following sources?").
Baiic census procedures involved the use of (he mail-
out/mail-back method for areas of the country contain-
ing 95.5 percent of the population and the conventional
method (i.e., door-to-door canvassing) for the remain-
der of the country. {See figure 19.2 for mail-census and
conventional areas.). This was essentially the same
approach as was used in 1970, except that the mail-out/
mail-back census procedure was employed more exten-
sively in 1980.
Recruitment and Training
District managers for the centralized district offices were recruited from among Census Bureau headquar-
ters personnel, and most of the RCC staff were persons
who had previous experience in the bureau's census
or survey work. All other personnel were temporary
employees hired only for the census. The manager and
key supervisors in the district offices were hired by the
RCCs, but other district office staff were hired at the
district office level.
The bureau's recruitment objective was to have the
127
staff in each district office representative of the popula-
tion of the area it covered. This meant that goals were
set for employing certain numbers of women and
minorities—blacks, Hispanics, Asians and Pacific islan-
ders, native Americans—or persons with proficiency in
certain languages, such as Spanish, Chinese, or Portu-
guese. If ws believed that a representative staff would
achieve the ' est count. Along these same lines, efforts
were made t . employ enumerators who lived in the enu-
meration dignéis (hey covered.
To meet the bureau's hiring ofcjeci'ves, various
sources were contacted, including elected officials, pub-
lic employment services, unemployment offices, and
community t=ased minority organizations. In addition,
free recruitment advertising was obtained from televi-
sión, i adío, and the newspapers; paid advertising was
used only when hiring goals could not be met through
other means
Regardless of the recruitment source, all job candi-
dates had to pass a written test and meet a minimum of
,er qualifications (citizenship was not required)
ucfore they could be hired. There were two versions of
the test, one for higher-level supervisors and one for
Figure 19.2. 1980 U.S. Census Mail and Non-Mail Areas
U S DEPARTMENT OF COMMERCE BUREAU OF THE CENSUS
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
tc: A FOSDIC-proof-'-iiíc, the clerks also looked at the
OaL^iionnaiuv for stray marks or written answers where
ther? should have been filled circles; the clrrks
transcribed the information from damaged question-
naires onto new ones. This questionnaire-edit operation
was conducted by placing a cardboard template over the questionnaires; the template had printed instructions
for the clerks to follow and a pattern of openings that
allowed answers on the questionnaire to show through.
Specific rules, differing by centralized and decentralized
offices, were applied to determine whether a question-
naire was within tolerance or failed the edit. A quality-
control operation was conducted to ensure the accuracy
of the edit.
For questionnaires that failed the edit check, an
attempt was made to resolve the pioblem first by
telephoning the household in question. Most of the
failed-edit questions were resolved in this way. Respon-
deros had been instructed to write their telephone num-
bers on the hacks of the questionnaires. The telephone
foliow-up was conducted by clerks in the district offices
in centralized areas, or from the enumerators' homes in
decentralized areas. In the cases where the problems
could not be resolved by telephone, enumerators paid
personal visits to the households involved. There was a
separate quality-control operation for the telephone fol-
low-up.
At tne completion of the first phase of field follow-up
ard the questionnaire edit, the second phase of field fol-
low-up began. The starting dates varied by district
office, depending on when the previous operations were
completed. Various kinds of cleanup work were con-
ducted as part of the second follow-up. The cases from
the first follow-up, where there was still no question-
naire for a household, were to be completed. The failed-
edit questionnaires requiring personal visit were also
pan of the work load in this operation.
The second follow-up also included one of the cen-
sus's major coverage-improvement checks, called the
vacant and delete check. All housing units classified as
vacant by the enumerators in the first follow-up and
most addresses deleted from the address registers in ear-
lier operations were checked by different enumerators in
the second follow-up to determine whether the units
w»re in fact vacant or should have been deleted. As a
result of this check, some units that had bten classified
as vacant were found to have been occupied, and some
addresses deleted from the registers were found to be
existing vacant or occupied units. Questionnaires were
cci-.pieted for the persons and housing units found in
earn case.
Enumeration in Conventional Areas
Cc iver.tional offices ; ere officially opened on 2.< Janu-
ary 1980. As in mail-census areas, muc'< of the activity
prior to census day centered around setting up th^
128
office, organizing supplies, and hiring staff. There were
no adJrosi-Iisune operations prior to census day a
there were in the mai!-ccn>us areas. Advance contacts
were made with the largest special places, and mailin,
pieces explaining the upcoming enumeration were sent
to those not contacted personally.
Postal carriers delivered unaddressed short-form
questionnaires to each known housing unit on 28
March. Householders were instructed to fill out the
form and hold it until an enumerator came to pick it up.
Beginning on 31 March, the enumerators canvassed
their assigned EDs, listing the address of each housing
unit, and collecting questionnaires or filling them out,
as necessary. At selected households, the enumerator
collected information on the long-form questionnaire
(the sampling pattern was preprinted in the enumera-
tor's address register).
A supplementary, American Indian questionnaire was
completed for every housing unit on Indian reservations
that was designated as a short-form housing unit and
had at least one American Indian, Eskimo, or Aleut
occupant. The supplementary questionnaire was also
used in mail-census districts where there were Indian
reservations and in parts of Oklahoma that were for-
merly Indian reservations (but not those in urbanized
areas).
A coverage check was conducted as a quality control
on the work of the enumerators. In advance of the cen-
sus, crew leaders made listings of 24 addresses in each
ED. After the enumeration was completed, the advance
listings were matched to the listing of housing units
made by the enumerator to determine whether the enu-
merator missed any addresses. If no address was missed,
the work was considered to be of good quality; if one
address was missed, the work was acceptable but the
address was added to the address register; if more than
one address was missed, the ED was recanvassed. Ques-
tionnaires were filled out for households and housing
units not previously enumerated.
In addition to the coverage check, the U.S. Postal
Service was used- to help improve census coverage in
conventional areas in an operation called the postenu-
meration post office check. Enumerators filled out an
address card for each housing unit they listed. The cards
were delivered to the post offices where carriers
reviewed them and noted addresses to which mail was
delivered but for which theie was no card. Theie
addresses were added to the address registers, if thev
were not already listed, and the units were later enu-
merated. The Postal Ser.ice also noted instances where
there was a card ic. a nonexistent addict. 1 :se
addresses were deleted fivm the. address registers.
Since all c u^tionnaires were returned to or filled out
by an cnumvv . -'.hos wor?: v ;s in turn checked b- •>
crew leader, i "a. not neeesrary *o have a fuli quest:
nai.e-edit o] ~ra*.i t: in the rui'-census offices.
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
129
'ion. Addiesses in these arcas also underwent the two
postal checks in March 1980.
On 28 March 1980, postal carriers delivered an
addressed questionnaire mailing package to every hous-
ing unit on the bureau's mailing lists. In addition to
either a short- or long-form questionnaire, the mailing
package included an instruction booklet and a return
envelope. Householders were instructed to fill out the
form and mail it back to the local census office on 1
April, census day. Return postage was prepaid (the Cen-
sus Bureau reimbursed the Postal Service for the costs
of the mail out and returns). The address of the district
office was printed on a label attached to the question-
naire; this label also contained important geographic
codes and other information.
To help respondents fill out their forms, the buieau
set up telephone assistance lines in each district office;
in most cases, these were toll-free numbers. The tele- phone assistance number for each respective district
office was printed on the questionnaire label, published
in local newspapers, and announced over television and
radio. It was estimated that more than 1.8 million calls
were answered. Walk-in assistance centers were set up in
some large cities in space donated by local community
groups. In addition, the community services specialists
and other bureau personnel conducted community
workshops on filling census questionnaires in the weeks
before census day.
As the questionnaires were returned to a district
office, they were sorted by ED and filed in a highly
secure part of the office to ensure the confidentiality of
census information. Only those census employees who
needed to work with a questionnaire were giv.-n access.
Persons who were not census employees were escorted
at all times when inside the office. Sucn per sor> were
not allowed (even with an escort) in any part of the
office where questionnaires were kept or piccessed.
After being sorted by ED, the questionnaire were ready
to be matched to the address register for that ED, an
operation called check in.
In most cases, the enumeration of special places
began on census day, but in some instances it began
earlier; for example, if a college recessed around census
day, the enumeration started a week or tv/c be!ore. Spe-
cial places (except for regular housing units within
them) were not enumerated by the maii-ccnsus method
but through various other means- -such as direct enu-
meration and questionnaire drop-off—depending on
the nature of the place. The special-places operation col-
lected information on persons living in college domino
ríes, prisons, hospital chronic wards, some nursing
homes, and other group quarters, and,, at hotels,
motels, missions, and street corners, enumerated trav-
elers who had no one at their usual home to count them, other transients, and persons with no usual place of resi -
dence As part of the advance work the district offices
mailed out posters explaining that the publicity asking
persons to mail back their questionnaires did not apply to those living m group quarters.
Two weeks were allowed for the receipt and check-in
of mail-return questionnaires, forms were returned to
the district offices for about 83 3 percent of all occupied
housing units. The first phase of follow-up began on 15
April and lasted six to eight weeks in most district
offices, though it took longer to complete in hard-to-
enumerate areas. A copy of the address register showing which units had and had not been accounted for was
given to enumerators who were to follow up on the non-
returns.
The follow-up work load consisted of nonresponse
units -housing units for which no questionnaire was
received. It was not known whether a nonresponse unit was occupied or vacant until a determination was made
by the enumerator. Enumerators went to each nonre-
sponse housing unit and either picked up a question-
naire if the householder had already filled it out or com-
pleted a questionnaire if the householder had not.
Enumerators also answered the housing questions for
vacant units. The enumerator's instructions encouraged
maximum self-response or self-enumeration by the
respondent. For instance, during an interview, the enu-
merator was instructed not to answer any item by obser-
vation, but to wait for a reply to each question from the
respondent.
The enumerators were expected ic work during the
hours when most people would be a: home, but not
before 9 a.m. or after 9 p.m. If the enumerator was una-
ble to find anyone at home after four visits, he or she
attempted to complete the questionnaire by observation
or by talking to neighbois, landlords, and building
superintendents—a procedure termed last resort. Every
effort was made to obtain at least last-resort informa-
tion. For population questions, last resort required that
the name of each person be collected along with three of
the following four characteristics: relationship to the
person in column 1 on the questionnaire, sex, race, and
marital status. Last resort also required answers to a
number of housing questions for both occupied and
vacant unitv Crew leaders were responsible fot meeting frequently
with their enumerators to pick up completed question-
naires, answer questions, and complete administrative
forms They used checklists to review the work of their
enumerators. An "assignment control" section in the
district office reviewed the enumerator-returned ques-
tionnaires for completeness.
The questionnaires returned by mail (and enumera-
tor-returned questionnaires in centralized offices) were edited by district office clerks to make certain that they
had been completed in an acceptable manner. To facili-
UN1T ED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
130
quality-control measure, a sample of the questionnaires
for each ED was reviewed for completeness by office
clerks. If the sample failed the review, the office clerks
edited all the questionnaires to identify those with miss-
ing information that should be included in the follow-up
operation.
Experience from previous censuses and tests indicated that enumerators dojiot always adhere to the sampling
pattern prescribed for collecting information on the
long form. Since many of the estimates from census
data are based on the additional information obtained
from long-form questionnaires, it is important that the
sample be accurate and representative of the total popu-
lation. After the enumeration, office clerks conducted a
sample tolerance check to see if the sampling pattern
had been properly employed. This was' done by com-
paring the actual population in an ED to an estimate
based on the number of people enumerated on the long-
form questionnaires for the ED. If the difference was
significant, the ED was "resampled" by transcribing
some long forms to short forms and vice versa; addi-
tional long-form information was collected where neces-
sary.
The follow-up in conventional offices was similar to
the second phase of follow-up in mail offices. Enumera-
tors telephoned or visited housing units to obtain the
required information on questionnaires that failed edit
or for refusal cases from the regular enumeration. EDs
that failed the sample tolerance check were resampled as
noted above, and EDs that failed the coverage check
were recanvassed. New addresses from the postenu-
meration post office check were visited and enumerated,
and, finally, the vacant and delete check was conducted
in the same way as in mail-census areas.
Local Review
The Local Review Program was one of several efforts
specially designed by the Census Bureau to improve the
completeness and accuracy of the 1980 census. It was a
part of census operations in both mail and conventional
areas. Its purpose was to allow local government offi-
cials in some 39,000 jurisdictions the opportunity to review the counts before they became final and while
census district offices were open and able to check
reported discrepancies. In addition, reviews were con-
ducted at about the same time by bureau staff in the dis-
trict offices and at headquarters. (After the district
offices closed, count complaints were handled by a unit
at bureau headquarters.) Local review pinpointed such
major problems as clusters of missed housing units, geo-
graphic misallocations, and incorrect geographic boun-
daries.
Field counts tallied in the district offices after the
completion of the regular enumeration or first follow-
up were sent to local officials who had 10 working days
to review the figures. Materials explaining the Local
Review Program had been sent to local officials in the
fall of 1979, and census maps had been sent to the loca-
lities several weeks prior to mailing out the field counts so that officials could organize their information. The
counts, which were provided for EDs and higher-level
geography, included total population, the number of
persons living in group quarters, and housing unit
counts, including the number and percentage of vacant
units. At this stage of the census, the housing unit
counts were more complete than the population counts,
and the officials were asked to concentrate their review
on the former.
If the local officials provided sufficient evidence of
discrepancies in the census housing unit counts, the dis-
trict office took several steps to check them, including a review of address registers in the office, a field recan-
vass of the area in question, or both. After receipt of the
local responses to the numbers, the district offices had
about four weeks to take action before they closed.
Closing the District Offices
After the completion of all follow-up work, a final man-
ual tally of population and housing counts was made in
each district office. The totals, referred to as prelimi-
nary counts, were reviewed by headquarters staff in
Washington. If the counts were acceptable, the district
office was authorized to close.
Most district offices closed four to six weeks behind
their scheduled dates, which ranged from II July for
conventional offices to 8 August for centralized offices.
The first district office closed in late July 1980, and
about one-fourth of the offices closed in August. By the
end of September, over 90 percent of the offices had
closed. The last offices to close were in large urban
areas, chiefly New York City. The final one was in the
Bedford-Stuyvesant area of Brooklyn, where an Octo-
ber fire destroyed most of the completed questionnaires
just before the office was to close. A reenumeration of
the area was completed m December 1980. Fire had also
struck the Framingham, Massachusetts, office near the
end of the first phase of follow-up, destroying many
of the questionnaires. A partial reenumeration was required, but the office was still among the earliest to
close.
In part because the number of persons and housing
units significantly exceeded precensus estimates, work
related to preparing for and completing the field enu-
meration exceeded the budgeted amount, forcing bud-
getary cuts or slowdowns in other phases of the census
(processing and publications).
Data Processing
As each district office closed, its questionnaires and
address registers were boxed and shipped to one of the
three processing centers—Jeffersonville, New Orleans,
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
or l aguna N.tuel. Their mission was io transfer the
dai.i from more than 90 rn:l!:or, questionnaires- ore for
etch housing ur.it plu^ the continuation question1": ires
for households with more than seven persons and ques-
tionnaires for persons in group quarters—onto com-
puter tapes so that they could be tabulated and cross
classified. This work required a great deal of space for
storage, temporary clerical staff, and sophisticated elec-
tronic equipment. The problems of controlling the flow
of questionnaires in one location (Jeffersonville) in 1970
led to the decision to distribute the clerical processing
among three sites in 1980.
A Decennial Processing Staff was created to organize
and control the clerical and precomputer work; the
actual computer processing was done at headquarters.
A processing center manager was responsible for the
overall direction of the operations in each of the three
sites and was aided by assistant managers for adminis-
tration and for operations. Top supervisors were, for
the most part, selected from among the bureau's career
staff, and clerical employees were chosen through the
civil service system.
Videotaped modules prepared by headquarters s! :Ji
were the primary training tools rather than the verbatim
guides followed in the field district offices, although
guides were used in some instances in conjunction with
audio-visual presentations.
Each of the three processing centers received millions
of questionnaires, thousands of address registers, and
vast quantities of other records. To maintain control
over these items, an automated inventory apd control
system was created. As the questionnaires, boxed by
ED, arrived at the piocessing centers, they were checked
in and given bar-code labels similar to those seen on
American grocery products. Each ED box had a unique
bar code. During processing, the labels were electron I
cally scanned as the boxes were checked in and out of
each work station. A manual control operation was
used as a backup in case of failures L ,.;-e automated
system. Thus, it was possible to teli where (he materials
for a given ED were at any tune.
Another control feature was the questionnaire and
address register library where materials were stored
when not at a work station. Materials were always
checked back into the library from one naioi operation
before being routed to another.
Equipment
The census questionnaires were specially designed to be
read by FOSDIC (film optical sensing device for input
to computers). Most ar.sv.ers on the qut.stionn.¡:; v.tre
filled circles, and any written response:; wire translated
into filled circles by coding clerks. During processing,
the questionnaires were photographed (and simultane-
ously reduced in size) on 16-mm microfilm, by hich-
131
suited camera units; there was an average of 20 camera
units at each site. The film was then developed and sub-
jected to various development and density checks.
The negative microfilm, on which ihe blackened cir
cles on the questionnaires appea-cd as ciear dots, was
scanned by FOSDIC. A tiny beam of light examined
each frame of microfilm, and where it found a clear dot,
the light beam initiated the production of magnetic
codes on computer tape. The meaning of the dots, in
terms of data, was interpreted by their position on the
microfilm. Names of individuals were not coded, and
did not appear on computer tapes during data process-
ing. The system was much quicker, cheaper, and more
accurate than keying the questionnaire data onto punch
cards. In a 24-hour period, one 1980 model FOSDIC
machine could transmit the information from 275,000
short-form questionnaires.
In 1970, the microfilm was flown4 to Suitland for
FOSDIC processing, but in 1980 four FOSDIC
machines were on site at each processing center. FOS-
DIC transmitted data by secure electronic means to the
main computer unit—the UNI VAC 1140—in Suitland
where it was stored on detailed basic-record tapes. The
output from these tapes, after further processing, was
used to produce all 1980 census products.
Data Preparation and Coding
Processing of the census questionnaires and the resul-
tant data occurred in two separate phases. The first
involved the 100 pet cent questions found on the short
forms and the initial pages of the long forms. The sec-
ond phar.e involved the additional questions on the
long-foim questionnaires and began when most of the
work on ihe 100 percent data was complete.
The i00 Percent Questions
The 100 percent data were given priority because they
included the final official state population counts that
the Census Bureau was required to provide the president
by 1 January 1981. In addition, the 100 percent data had
to be processed firs- in order to provide (I) the analyti-
cal tools fo; evaluat ing the accuracy of the data and (2)
the weighting controls for inflating the sample responses
to reflect the total population. The 100 percent process-
ing could be done quicker because it did not involve
the time-con." urning hand coding required of certain
responses on the sample forms.
Materials from ihe firs? district offices to close
arrived ai the processing centers during the first weeks
oí August 1980 were checked in, and were prepared for
filming The fu;-? data were tran:mitten via FOSDIC to
•¿te cur;¡putf. on 7 August. The flow of questionnaires
from the di.-irict offices was slow at first, then built up
ihiougb ear'v October. The proccvsinp, centers had
'iff;culty keeping up with tiv Ft w, th 'ate fi:'
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
132
quality-control measure, a sample of the questionnaires
for each ED was reviewed for completeness by office
clerks. If the sample failed the review, the office clerks
edited all the questionnaires to identify those with miss-
ing information that should be included in the follow-up
operation.
Experience from previous censuses and tests indicated that enumerators dojiot always adhere to the sampling
pattern prescribed for collecting information on the
long form. Since many of the estimates from census
data are based on the additional information obtained
from long-form questionnaires, it is important that the
sample be accurate and representative of the total popu-
lation. After the enumeration, office clerks conducted a
sample tolerance check to see if the sampling pattern
had been properly employed. This was done by com-
paring the actual population in an ED to an estimate
based on the number of people enumerated on the long-
form questionnaires for the ED. If the difference was
significant, the ED was "resampled" by transcribing
some long forms to short forms and vice versa; addi-
tional long-form information was collected where neces-
sary.
The follow-up in conventional offices was similar to
the second phase of follow-up in mail offices. Enumera-
tors telephoned or visited housing units to obtain the
required information on questionnaires that failed edit or for refusal cases from the regular enumeration. EDs
that failed the sample tolerance check were resampled as
noted above, and EDs that failed the coverage check
were recanvassed. New addresses from the postenu-
meration post office check were visited and enumerated,
and, finally, the vacant and delete check was conducted
in the same way as in mail-census areas.
Local Review
The Local Review Program was one of several efforts
specially designed by the Census Bureau to improve the
completeness and accuracy of the 1980 census. It was a part of census operations in both mail and conventional
areas. Its purpose was to allow local government offi-
cials in some 39,000 jurisdictions the opportunity to review the counts before they became final and while
census district offices were open and able to check
reported discrepancies. In addition, reviews were con-
ducted at about the same time by bureau staff in the dis-
trict offices and at headquarters. (After the district
offices closed, count complaints were handled by a unit
at bureau headquarters.) Local review pinpointed such
major problems as clusters of missed housing units, geo-
graphic misallocations, and incorrect geographic boun-
daries.
Field counts tallied in the district offices after the
completion of the regular enumeration or first follow- up were sent to local officials who had 10 working days
to review the figures. Materials explaining the Local
Review Program had been sent to local officials in the
fall of 1979, and census maps had been sent to the loca-
lities several weeks prior to mailing out the field counts
so that officials could organize their information. The
counts, which were provided for EDs and higher-level
geography, included total population, the number of
persons living in group quarters, and housing unit
counts, including the number and percentage of vacant
units. At this stage of the census, the housing unit
counts were more complete than the population counts,
and the officials were asked to concentrate their review
on the former.
If the local officials provided sufficient evidence of
discrepancies in the census housing unit counts, the dis-
trict office look several steps to check them, including a review of address registers in the office, a field recan-
vass of the area in question, or both. After receipt of the
local responses to the numbers, the district offices had
about four weeks to take action before they closed.
Closing the District Offices
After the completion of all follow-up work, a final man-
ual tally of population and housing counts was made in
each district office. The totals, referred to as prelimi-
nary counts, were reviewed by headquarters staff in
Washington. If the counts were acceptable, the district
office was authorized to close. Most district offices closed four to six weeks behind
their scheduled dates, which ranged from 11 July for
conventional offices to 8 August for centralized offices.
The first district office closed in late July 1980, and
about one-fourth of the offices closed in August. By the
end of September, over 90 percent of the offices had
closed. The last offices to close were in large urban
areas, chiefly New York City. The final one was in the
Bedford-Stuyvesant area of Brooklyn, where an Octo-
ber fire destroyed most of the completed questionnaires
just before the office v. as to close. A reenumeration of
the area was completed in December 1980. Fire had also
struck the Framingham, Massachusetts, office near the
end of the first phase of follow-up, destroying many
of the questionnaires. A partial reenumeration was required, but the office was still among the earliest to
close.
In part because the number of persons and housing
units significantly exceeded precensus estimates, work related to preparing for and completing the field enu-
meration exceeded the budgeted amount, forcing bud-
getary cuts or slowdowns in other phases of the census
(processing and publications).
Data Processing
As each district office closed, its questionnaires and
address registers were boxed and shipped to one of the
three processing centers—Jeffersonville, New Orleans,
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
133
reponed" categcicf we: c added to each tabulation,
r^riain printed -cpuran.i most summary tape files
i- .¡jded tables .bowing (he airomt of allocution and rubs!itutk .1 for certain ¡reins.
Following computer editing, the sample data went
through a procedure that assigned a wci¿ht to each sam-
ple person and housing unit. In areas sampled at the rate
of 1 -in-2, the sample weights were close to two. In areas
sampled at the rate of 1 in-6, the weights averaged
about six. The weights were assigned in such a fashion
íh¿t for most large geographic areas the 100 percent cen-
sus counts and the sample tabulations for total popula-
tion and total housing units were very close.
In addition to the ED-level data check done in the
d._ry review, the data for larger areas—e.g., states,
SMS As, counties, and MCDs—were checked through
elaborate computer programs called anal>/ers. The pur-
pose of the analyzers was to assure that the statistics-for
larger areas conformed to their expected levels based on
the previous census or intercensal surveys. The analy-
zers also were used to check the totals for population
and housing unit counts and to monitor the allocation
rates of various population and housing characteristics.
Confidentiality of Census Records
Once the above processes were completed, edited data
about individuals and housing units, together with asso-
ciated geographic information, were stored on basic
record tapes (BRTs). All 100 percent and sample tabula-
tions were made from these tapes. Although the BRTs
do not contain names and addresses, they do have
detailed geographic codes and household data that
could result in the disclosure of data for individuals;
therefore, these tapes are confidential and may be used
only by bureau employees in preparing statistical piod
ucts.
The original questionnaires were destroyed and the
pulp recycled about the time the processing centers
closed, although a small percentage of 'he question
naires were retained longer for processing and evalua-
tion purposes. One microfilm copy of the question-
naires was placed in the custody of the National
Archives, and another was retained for use in age-search
processing at the bureau's Pktsburg, Kansas, facility.
Products and Dissemination
Once the data were entered on the brsic record tapes,
Lhe production of census data could begin. The primas y
product was a series of five summary Uoe files f.STFs)
Thes? computer tapes provided ci.iy tabulation or fre-
quency distributions and not ;.ae individual records (which were on the basic tape:) i\u thus could i:c sold
for public use. The STFs were u;ed to produce i_ts of
t'bul3tior.r mat appeared in printed cepo its and on
r..'; re fiche The printed rf.f oits served a broader audi-
ence than the SIFs, but were more costly to produce
and contained onl> a ¿mall portion of the tabulation on
the STFs.
Budgetary constraint required th:it some economies
be made in the original data release plans. For instance,
one report series—Block Statistics—originally intended
to be issued as printed reports was issued only on micro-
fiche, a less expensive process. The bureau had planned
to produce microfiche of all paper reports, but, in order
to save money, the reports were generally issued either in
print or on microfiche, but not both. Microfiche copy
was issued only for STFs 1A and ?A rather than for sev-
eral STFs as originally planned, and plans to combine
the individual paperbound reports for some of the pop-
ulation and housing scries into hardbound volumes were
dropped.
+ The bureau's new policy for correcting the counts
contributed to a delay in issuing the 100 percent data
products. In 1970, when population and housing count
errors—resulting from geographic misallocations or
processing operations—were detected, the bureau used
errata sheets in the printed volumes to notify users of
corrections. While such sheets were still used to some
extent in 1980, wherever possible corrections were made
in the basic record tapes from which all tabulations were
generated.
The first data released in the 1980 census were the pre-
liminary housing unit and population totals provided
when the district offices closed. Each locality was sent a
mailgram announcing its preliminary counts, and the
bureau simultaneously issued these figures in press
re,case) The announcement, of the held counts for some
39,000 loca! governments was completed in mid-Decem-
ber 1980.
The first computer generated counts were released at
the end of the 100 percent data processing. These were
the official state population totals, transmitted by the
director of the Census Bureau to the president on 31
December 1980 along with the number of seats to which
each state was entitled in the House of Representatives.
The daia thai the bureau was required to deliver
under Public Law 94--171 to state apportionment and
redistneting offices by 1 April 198! were released in
February and March on computer tape, microfiche, and
eye-readable papa prints from the microfiche. These
data files contained figures for total population and
provisional figures for persons of Spanish origin end for
five race groups, white; black; American Indian,
Eskimo, and Aleut; Asian and Pacific islander; and
"Other" The data v cv;'.own foi the lowest geographic
;v.eas —b'..cks, or election prec nets in certain
states Maps were also provided to the states.
Summary Tap;"s
As mentioned f ■ wic .siy, the rru jci )on;>n cf 'he results of the 19'iO censu. wen province in a seí ; í fr ?
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
134
office closings threatened efforts to meet the 1 January deadline for transmitting final state population counts
to the president. Materials for the Bedford-Stuyvesant
office in New York arrived at Jeffersonville just days
before the deadline, but all were processed, and the 100
percent clerical and electronic operations were comple-
ted on time.
The Sample Questions
After completion of the clerical processing of the 100
percent data, the long-form (sample) questionnaires
were separated from the short forms and readied for
hand coding. Some of the census questions (e.g., ances-
try, language, industry, occupation, place of work, and income) required a written answer from the respon-
dents, and coding was essential to convert these written
answers into a machine-readable form. Most of the
questions requiring coding (25 population and 7 housing
items) were on the sample pages of the long form; how-
ever, two questions on the 100 percent pages of the long
form—race and relationship—had to be coded when-
ever the respondents wrote in an answer rather than fill- ing one of the circles. Where the written entry could
be assigned to one of the circles already on the ques-
tionnaire (e.g., assigning a written "Caucasian" to
"white"), this was done in the district offices. Numeri-
cal coding of these write-in entries was done only for
sample questionnaires in the processing offices.
The coding operations were conducted by three sepa-
rate sections of clerks: one section worked solely on the
place-of-work, travel-time-to-work, and migration
questions; a second on the industry and occupation
questions; and the third on all other items. A quality-
control operation checked the clerks' work. Once the
sample questionnaires for a group of EDs were com-
pletely coded, they were filmed and processed via FOS-
DIC in the same way as the short-form questionnaires.
Because of budget constraints, a decision was made in
February 1981 to reduce the staff, slow down the sample
coding operation, and stretch the work out into the next
fiscal year (i.e., beyond October 1981). At the same
time, it was decided that the place-of-work, travel-time,
and migration questions would be coded on only half of
the sample questionnaires, thus reducing the sampling
rate for these three items. When the budget situation
improved in June 1981 with the appropriation of sup-
plemental funds, the bureau stepped up its coding oper-
ation by increasing staff; however, the sample reduction
for the three items mentioned above was retained. Cod-
ing, which had begun at all three sites during January
1981, was completed by the end of October. Filming and
FOSDIC transmission was accomplished between July and December. The Laguna Niguel and New Orleans
centers and decennial census operations at Jeffersonville
closed down in early 1982.
Editing and Imputation
After the 100 percent data were put on computer tape,
and again after the sample data were entered, the counts
for each ED had to pass a set of acceptance tests to
ascertain that data scanned by FOSDIC had not been
lost or incorrectly recorded on tape, and that potential
errors or unusual entries did not exceed established tol- erances for population, housing units, and various pop-
ulation and housing characteristics. Essentially, this
involved comparing the initial computer counts to the
1980 field counts. The data that failed were summarized
and printed out for each ED in a format called a diary.
To clear up problems, clerks in the processing centers
compared the data on the ED diary to the boxes of ques-
tionnaires, the address registers, and the microfilm to
'make sure all the forms were in good shape (not crum-
pled, markings sufficiently dark, etc.), properly identi-
fied, and that none were missed during the filming pro-
cess. When necessary, rejected EDs were remicrofilmed
and sent through FOSDIC once again.
A series of computer edits was employed to fill m
missing data on the questionnaires or to account for
inconsistent data. Despite the fact that every effort was
made to complete the questionnaires in the field, some
questionnaires were returned with items not completed and some corresponded to "unclassified" housing units
—addresses for which the occupancy status could not be
determined in the field. Responses were edited to elimi-
nate inconsistent information based on data from a given record. For example, if the reference person was a
married male and the marital status of the female
recorded as his wife was "divorced," then the latter's
marital status was changed to "married."
A procedure called allocation was used when missing
or inconsistent information could not be supplied or
corrected on the basis of other entries on the same
record. When this happened, the computer selected a
reasonably r tched housing unit and imputed (or allo-
cated) its characteristics for the missing or inconsistent
information. A similar procedure was used for missing
population characteristics: data were imputed by refer-
ring back to the previous records for a household or per - son with characteristics resembling the ones for which
data were needed.
In addition to allocation, substitution was sometimes
used. This occurred when a person or housing unit was
known to be present, but no characteristics were
recorded. In this case, data from a previously processed
housing unit were selected as a substitute and a full set
of characteristics for the housing unit and for each per-
son in the unit was duplicated.
The computer editing procedures were designed to
make the census statistics a more useful description of
the nation's population and housing than if "not
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
135
<ra:e down to places of i .000 or rr.orc people; includes
c:c- hy i ;>cc anj Spanish origin, lss.ied
: r:\iccn Apiii amiN 'wmber !9S2.
PC^O i -C, Cerwu! Socul j;tü Econvr,¡c Charac'.cris-
; Vs. Sample population subjects for places of 2,500 or
more inhabitants and higher geographic levels. Issued
between July and November 1983.
PC80 1 -D, Detailed Population Characteristics.
Issued on microfiche only. This series shows most sam-
pe population subjects, cross classified by age, race,
ar.d other characteristics, for states and for SMSAs of
If0,000 or more people. The first of these reports was
t>>u:d in September 1983.
PCS0-2, Subject Reports. Detailed data on various
population characteristics, principally at the national
a%d regional levels.
19S0 Census of Housing
HCS0-1-A, Genera! Housing Characteristics. Data for
the complete-count housing items tabulated from the
state level down to places of 1,000 inhabitants; includes
cross-tabulations by the householders' race or Spanish
origin. Issued between July and November 1982.
HC80-1-B, Detailed Housing Characteristics. Sample
housing items tabulated down to the level of places with
2,500 inhabitants; includes some tabulations by race
and Spanish origin of householders. Issued between
July and October 1983.
HC80-2, Metropolitan Housing Characteristics. This
series presents most of the 1980 housing subjects in con-
siderable detail and cross-tabulation, with separate
reports for each state as well as each SMSA. Issued
between November 1983 and February 1984.
HC80-3, Subject Reports. Data on detailed topics at
the national and regional level.
HC80-4. Two reports for the Components of Inven-
tory Change (CINCH) Survey. CINCH data were shown
for the nation, regions, and size groupings of SMSAs.
HC80-5. Report for the Residential Finance (RF) Sur-
vey. RF data were produced for the nation, regions, and
size groupings of places.
Other Products
Several series of supplementary reports were issued in
¡r'rted form as a means of disseminating selected popu-
\l::on and housing data for larger geographic areas in
advance of the regular reports. The first supplement
issued, in May 1981, was Age, Sex, Race, and Spanish
Origin of the Population by Regions, Divisions, and
Sietes: /980, PC80-S1-1.
Another of the supplementary reports was Provi-
sional Estimates of Social Economic, and Housing
~ >cractcristics, PHC80-S2-L I:.sued X April 198?, it
.•jitained the fit sample tabulations ¡rom the census,
i j was produced to compensate for delays in the regu-
len "ample data pioducts. The report was ba>ed on .he
long forms in a sample of about 17,000 FDs. v hich con-
tained slightly more than 1.5 peicent of the housing
units and persons in the nation. Da:c w:a- proviJ^d fcr
the nation, the states, the District of Lo umbia, and the
38 SMSAs with 1 million or more inhabitants.
To further compensate for the late release of sample
data, the supplementary reports, Advance Estimates
Of Social, Economic, and Housing Chcracterisncs,
PHC80-S2, were prepared. There was one report for
each state, the District of Columbia, Puerto Rico, and
each outlying area, in a format and with detail similar to
that in PHCS0-S1-1, and giving data for the state, coun-
ties, and places of 25,000 or more persons. Unlike
PHC80-SI-1, this series of reports was based on a full
census sample.
A special computer file contained sample data useful
in equal employment opportunity and affirmative
action planning. The file contained all 503 occupation
categories recorded in ihe census by sex, race, and Span-
ish origin, for all counties and for cities of 50,000 or
more persons, plus certain tabulations relating to educa-
tional attainment by age.
While the bureau's STFs contained a wide selection of
data tabulations, they could not meet the specialized
needs of all data users. Thus, the bureau prepared pub-
lic-use microdata samples, containing actual census
records, to allow users to make their own special tabula-
tions for large areas. The primary consideration in
designing these files was to provide as much informa-
tion as possible while protecting individual confidential-
ity. No names or addresses were on the files since they
were not on the basic record tapes, and each geographic
area identified had to have at least 100,000 inhabitants
so that individuals could not be identified by the charac-
teristics given. There were three files, one containing 5
percent of the population and two with 1 percent each.
The bureau also had the capacity to do special tabula-
tions on a cost-reimbursable basis. While these were
more expensive for users than purchasing public-use
microdata files, the bureau did offer several standard-
ized tabulations more cheaply. One of these was issued
under the Neighborhood Statistics Program, in which
participating localities defined neighborhoods in terms
of census geography.
A computerized Master Area Reference File (MARF)
was issued for use with STFs and other tape products.
The MARF contained numeric codes and names (where
appropriate) of geographic areas, and selected popula-
tion and housing counts.
Census maps for 19-,0 wee made available to users as
they were completed. The¿e included five types of
detailed maps: county, place, place-and-vicinity (where
there were built-up an as a: jund the place), Indian-res-
ervation, and maps f -r the 'y .-vt;!ed portion; of metrcv'ch'an counties. Tries. dt,.n!ed maps that
included areas covered in the block caustics program
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
136
summary tapo files (STFs) for data users with access to
computer facilities. The STFs, which were released on a
state-by-statc basis, provided data with much grcate.
subject and geographic detail than was feasible or desir-
able 10 publish in printed reports.
The first two STFs related to the population and
housing subject items collected on a 100 percent basis.
STFs 3, 4, and 5 cpntained subject items collected on a
sample basis, and generally included sample data for the
100 percent items for purposes of cross classification.
The first four STFs had varying degrees of small-area
data, while the geography in the fifth was limited to
states, SMSAs, central cities of SMSAs, and counties
and places with 50,000 or more inhabitants. Most of the
STF series were divided into files with different geo-
graphic structures. For instance, STF 1A provided data
down to block group or ED level, but STF IB gave data
for blocks; STF 2A contained tract statistics, but in STF
2B the lowest geographic levels were places of 1,000 or
more and county subdivisions. The C files were national
in scope, with figures for higher-level entities such as
places with 10,000 or more inhabitants and counties.
The first summary tapes were released in August
1981, and all had been released by the end of 1983. For
both summary tapes and printed reports, the data for
stales with smaller populations were generally issued
first.
Printed Reports
The bureau's printed reports appeared in paperback
volumes or on microfiche and were released under three
subject titles, 1980 Census of Population and Housing,
1980 Census of Population, and 1980 Census of Hous-
ing. Each series generally contained a summary report
for the United States, and reports for each state, the
District of Columbia, Puerto Rico, and each outlying
area. These reports are described below. Issue dates are
Table 19.2 Relation of Selected Printed Reports to the
Summary Tape Files
Primed Report Source
PHC80-1 Block Statistics (Microfiche) STF IB PHC80-2 Census Tracts STF2A.4A PHC80-3 Summary Characteristics for
Governmental Units and SMSAs STF 1 A, 3A PHC80-4 Districts of the 98th Congress STF ID, 3D PC80-I-A Number of Inhabitants STF IA PC80-1-B General Population Characteristics STF2B PC80-1 -C General Social and Economic
Characteristics STF 4B PC80-1-D Detailed Population Characteristics STF 5 HC80-1-A General Housing Characteristics STF 2B HC80-1-B Detailed Housing Characteristics STF 4B HC80-2 Metropolitan Housing
Characteristics STF 5
for the reports for the states and the District of Colum-
bia.
1980 Census of Population and Housing
PHC80-P, Preliminary Population and Housing Unit
Counts. Produced from district office hand tallies and issued between November 1980 and February 1981. This
series shows preliminary population and housing unit
counts for the state, counties, county subdivisions,
incorporated places, SMSAs, and congressional dis-
tricts. PHC80-V, Final Population and Housing Unit
Counts. Issued between February and April 1981. This
series presents official computer-generated population
counts, classified by provisional race and Spanish-origin
data, and housing unit founts for the state, counties,
county subdivisions, incorporated places, and congres-
sional districts.
In the PHC80-P and PHC80-V series, press releases
only were issued for the Trust Territory of the Pacific
Islands and the Northern Mariana Islands.
PHC80-1, Block Statistics. One report for each
SMSA showing data for individual blocks on selected
100 percent subjects; issued on microfiche only. The
series also includes a report for each state and Puerto
Rico presenting block statistics for cities of 10,000 or more people outside SMSAs and for communities out-
side SMSAs that contracted with the Census Bureau to
provide block statistics. These reports were issued
between February and November 1982.
PHC80-2, Census Tracts. Statistics on most census
subjects by tract for each SMSA and for other tracted
counties in a remainder-of-state report. These reports
were issued between July and October 1983.
PHC80-3, Summary Characteristics for Governmen-
tal Units and Standard Metropolitan Statistical Areas.
Issued between October and December 1982. Selected
complete-count and sample population and housing
data for SM :>As and each of some 39,000 general-pur-
pose local governments in the United States. PHC80-4, Districts of the 98th Congress. Selected 100
percent and sample population and housing characteris-
tics for congressional districts drawn, as a result of post- 1980 census redistricting, for the 1982 elections.
Selected data are also presented for counties and places
with over 10,000 people. These reports were issued
between March and May 1983.
1980 Census of Population
PC80-1-A, Number of Inhabitants. Population counts
(for the 1980 and previous censuses) for ail counties,
county subdivisions, places, SCSAs, SMSAs, and urbanized areas, and by urban/rural residence, issued
between October 1981 and May 1982. PC80-1-B, General Population Characteristics. Com-
plete-count data for various geographic levels, from the
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
137
more extensive, probinc questions to measure thr con-
sistency and accuracy of icportm^.
The utility-cost record check study was conducted to evaluate the accuracy of the reporting of average
monthly gas and electricity costs (question H22). Expe-
rience from the 1970 census and 1980 census tests indi-
cated that respondents tended to report higher than
actual expenditures. In this study, a sample of half of
the utility customers in eight selected areas received a
statement from their utility companies just before cen-
sus day showing their average payments for the previous
12 months. It was believed that this information could
improve reporting of utility costs. The response errors
of the individuals who received the information uere compared to those of the individuals in these areas who
did not. The program also examined the data improve-
ments that could result from supplying utility-cost
information to respondents.
Experimental Programs
A number of experimental programs examined alterna-
tive approaches to the 1980 census procedures. Gener-
ally, each experiment was implemented in only a frac-
tion of the district offices. The alternative questionnaire
experiment, however, employed a national sample and
was not limited to specific district offices. The appropri-
ateness of conducting an experiment along with the cen-
sus was a major consideration in deciding which pro-
grams to test for 1980. Several procedures were tested,
some dealing with variations in the enumeration process
and others with alternative methods of recruiting, train-
ing, and motivating enumerators.
The update/list/leave experiment studied alternatives
to the delivery of questionnaires by the Postal Service.
The telephone follow-up of nonresponse experiment
explored the cost-effectiveness of following up nonre-
sponding households by telephone, rather than by per-
sonal visit. The alternative questionnaire experiment
tested the effect of questionnaire design on mail-return
and item-completion rates. The alternative training
experiment compared standard enumerator training
with an alternative method. The purpose of the job
enrichment experiment was to reinforce the individual
enumerator's motivation and job knowledge with extra
training and experience. The student intern program
tested the feasibility of recruiting and employing college
students as census takers, particularly in minority areas.
Other research and evaluation projects addressed
coding, imputation procedures, quality controls, tlto
publicity program, and various sources of error in cov
erage and content.
Litigation and Adjustment Ij suc
The 1980 census may well be the most litigious ever. Oi'e
of the major suits against the Census Bureau was filed
ovn bcforo the census began. In 1979 the Federation for
American Immigration Reform (FAIR), consisting of
n,ore than 100 persons throughout the country, sought
to require the Census Bureau to exclude illegal aliens
from the total population for state and federal reappor-
tionment purposes. The government argued that it was
unconstitutional to exclude from the census any persons living in the United States (except residents of embas-
sies, etc.). Furthermore, the 1980 census questionnaire,
which was being printed when the FAIR suit was intro-
duced, did not contain a question that asked whether
one was an illegal alien, and the bureau believed that the
inclusion of such an item would seriously hamper its
efforts to achieve a complete count by making illegal
aliens more reluctant than ever to respond to the census.
The FAIR suit was dismissed in February 1980 by a
lower federal court on the grounds that the plaintiffs
had no legal standing to bring the suit. In November
1980 the appellate court affirmed the lower court's deci-
sion, and in March 1981 the Supreme Court refused to
hear the case. In the meantime, apportionment counts
consisting of the entire resident population had been
delivered to the president.
By October 1981 about 50 suits had been filed against
the bureau by states, counties, cities, an American
Indian tribe, a special interest group, and private citi-
zens. Most of the cases concerned one or more of three
broad issues: (1) allegations of mismanagement or pro-
cedural inadequacy in the field activities; (2) access by
the plaintiffs to census materials, such as address regis-
ters, that contained confidential information; and (3)
the adjustment of census figures for undercount. In
January 1981 many of the lawsuits that had not yet had
final judgments rendered by a district court were conso-
lidated.
Support for adjustment of the census had been voiced
by several sources throughout the 1970s and gained
momentum with a study by the National Academy of
Science's Panel on Decennial Census Plans in 1978,
which concluded that adjustment was feasible. Begin-
ning in the late summer of 1979, the director of the Cen-
sus Bureau initiated a program designed to help the
bureau reach a decision on whether to adjust. The cen-
terpiece of the program was a conference on census
undercount, held in February 1980, and attended by
more than 140 academic, governmental, business, and
kga! professionals. Attendees considered alternative
approaches for measuring the undercount and assessed
the implications of adjustment. The primary task of the
co.!^re:jwe, and cf two bureau internal workshops held
in the fall of 1979 and the fall ol 1980, was to examine
the critical underlying assumptions that would establish
a proper framework for deciding whether, wnen, &■ d how to adjust the counts. Papers i'.-.m the conference
; -A the workshops were circJ j.ied to a wide audienr
"or comments. Thioughotn Luis pidie erects
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
138
■-ere published in conjunction with PHC80-1 (Block
Statistics). While the Government Printing Office issued
the block statistics reports on microfiche, it provided the maps in paper form; the Census Bureau planned to
sell the maps on microfiche later. In addition to these, a
number of maps were published with the reports. The
maps were not necessarily issued at the same time as the
reports.
Computer tape products and most maps were priced
and sold by the Census Bureau's Data User Services
Division. The major series of printed reports were
priced and sold by the Government Printing Office.
Generally, publications were issued free to Commerce
Department field and Census Bureau regional offices,
to over 1,300 government and census depository libra-
ries throughout the country, and to state data centers.
The state data centers also received summary tape files
for their areas. The state data center program is a fed-
eral-state cooperative effort in which state agencies and
their affiliates receive basic data products and training
so they can assist public agencies and private users. The program began in 1977 and by the end of 1983 extended
to 49 states, the District of Columbia, Puerto Rico, and the Virgin Islands.
The bureau also established the National Clearing-
house for Census Data Services—a group of private,
academic, and public organizations that offer data
retrieval and related services to outside customers. The
Census Bureau does not, however, regulate or endorse
any of the registrants.
Evaluation
Forma] evaluation has been an essential component of
each census since 1950. As part of the 1980 census, there
were two main areas of evaluation (coverage and con-
tent), several major experimental programs, and various
research projects.
Coverage Evaluation
The first objective of the coverage evaluation program
was to develop estimates of coverage in the 1980 census
for both population and housing units. The second
objective was to evaluate the special procedures and
operations designed to improve coverage in the 1980
census. There were two main programs used to estimate the
coverage of the population. The first was demographic
analysis, which aimed at providing national estimates of
net census error for age, sex, and race groups. This
involved combining various types of demographic data
from sources essentially independent of the census—
such as birth, death, and immigration records—and comparing the resulting estimates with census counts.
This method was limited by the lack of acceptable esti-
mates of the illegal alien population.
The second program was the Post-Enumeration Pro-
gram (PEP). The Census Bureau conducts a Current
Population Survey (CPS) each month to collect current
labor-force information and socioeconomic data from
the civilian noninstitutiona! population (the monthly
unemployment figures are produced from CPS data).
The CPS is jointly funded by the Census Bureau and the
Bureau of Labor Statistics. The PEP checked 150,000
households surveyed in the April and August 1980 CPSs
to see if the persons in these households had been enu-
merated in the census. Persons in the CPS who were not
found in the census were counted as "gross omissions."
This estimate of omissions was inflated because of geo-
graphic errors and deflated because of duplicate enu-
merations. To compensate for these factors a sample of
110,000 households from the census was reinterviewed
to see if residents had been included correctly in the cen-
sus, that is, to see that they had been enumerated only
once and in the proper geographic area. Results from
the CPS-census matching operation were combined with
the results from the reinterview sample to provide dual-
system (census and CPS) estimates of the population.
The PEP population estimates were compared with cen-
sus counts to estimate the net census undercount for the
nation by certain age, sex, race, and Spanish-origin cat-
egories, and for regions, states, and large SMSAs with-
out such demographic detail. The estimates from the
PEP are limited by problems in matching cases between
the census and the CPS or reinterview.
In addition to demographic analysis and the PEP,
which provided estimates of population coverage only,
there was a separate study of housing coverage. Although coverage-improvement techniques had been
employed in previous censuses, some new ones were
used in 1980 and some old procedures were enhanced.
Among those studied in the coverage evaluation pro-
gram were the vacant and delete check, the matching of
census records 'o driver's license lists, the postenumera-
tion post office check, and the questionnaire assistance
centers. The evaluations measured the improvement in coverage resulting from the operations relative to their
cost, and determined whether they had been imple-
mented correctly.
Content Evaluation
Content errors are any errors that result in an incorrect
classification of a population or housing characteristic.
The two main content evaluations for 1980 were a con-
tent reinterview study and a utility-cost record-check
evaluation. In the content reinterview, about 12,000
households enumerated on long-form questionnaires in the census were reinterviewed and the answers from the reinterview were compared with those given during the
census. The reinterview focused on items that were new
or substantially changed for 1980 (e.g., Spanish origin,
ancestry, English-language proficiency), and included
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
Piense fill out tins
official Census Form
3nd moil it back on
Census Day.
Tuesday. April 1. 1980
Your answers are confidential p, M.* ir' » ') ^ Co'iei «f» &uD*fci to ' r • >1 O' ^ 'o« a'-? <1 si insv'* c' vou' jfM<*•'S 0"'i *'if« f7 vfj t dn*t yOu' if-toihfr.i-nr in ' ' g ivrinn-fn) aC]tnC«f1 O' PuM C ' '•* *-"■* 1 J** ' p t,!'•.♦' »>> .1 j' Ihf Qu#v IO I*" t)*SI Of yOu' k rv.^if i1q#
P,irn personas de habla hispana <f.. «■{,»•• «,'V 1 o m M * i'*- (• i v*^r?to OK rfuso i ^ I^ANCM i-.iri., * ■ i nil, .*, % ,k i K) (i rním#H0 ii>* f'"w tie l' >»■ > »' « M» ',t i|h« <_W>n
» '%wqiir , a»*Ma O y rVwj#*v.í ** fuf,hor'*'*) v • L *n f n f' \ 'Nf .iL^" vr f% kifff
u S f'i»l l-l-'. 'M •* (
1980
Census of the
United States
A m*####* k#m fh# Immu of Ih# C*n#u#
W# mj»|. from tim# to l#%# iiorh o' Ou'M'v#i •• t CéOCi'« if Our H#l'On >1 to m#0T luCC#4S''»Hy fh# '«|Oy nil-onil
lor#l c ,* l#r,gA# w# l*C# ^|| I) ^•pu<POt< o' I OSO
Th# #gi#ntal r#md 'o« * popu'*t'0<> w#a '#cognt#d #lmo?i 700 y##' # «Q0 wh#n Ou' Cp'»«^lul»0n *a? wi<M#n A| p/ovd^d Cry |r l>c'a I ih« first *45 condu^ '»d ,n I )90 *nd &n« h*; fct#n t««|H «v«/y I 0 y# #« # t«r>c# lh#m Th# l#w uh<?#r wh<ch th» # <ahtn prottci» th# conl./i#' o* yo«i« ^o< th# n#*i 7? y#arj — or uniiiAp'l 1 70S? — onty iwoi n c •'•íl.i aoUi & (cotl 10 :h* r,(j,Yidu*i J#co'di rrvl nn on# «lift mi^ s#a lh#m YOüf #niw#i $ wh#fk w»fh th# jnjw#() from oth#f peop fl w»s! p'Ov.rl# ih# siaMt<c#> r.p»d#d by public grwj P'tvjt# qiowpi 3ChrK)f» t>u1'n#SJ And <ndu^1rv ^ Ffdr 4l SlJl« #rc] toCÍ' gOvpf-KM^nt^ »CPC<I Ih# COv.»nliy Th-,# t'Qu'f* «vil h#ip *ti i«clois of A^ri#fic#n socifliy unde'JTiind ho* Oi»« POpuUliun *rn' houl»riq #re changing In Ihii w#v vv« c*n df*' moii €l'#Li'v»ty wlh today 5 probiftf"1 amri wo'k luws'd a b*"#r 'utu'e >0' O1 ui The t#niu» >$ i wiu'i* impmiirn n*;,,jA*i »cl«vity Pl#ai# 6o four p#«l bv t'Umq ou' th 1 rentos tc«t> arcu'atpty an«J CO"ip'*t^fy II you m¿'' »| ha^ h | 'nmpHy ,n ih# enclosed 00S'»<4# o#'d #nve'op< vnJi %# # the ««pens# a*»d ' nc Anv#n,#nr # of 4 Cf^»u3 l»*P' hj ,nq fi vil t yQu fhjnh yi'Hj 'or you' Conp#'»t &n
Pl«i«l conliny* —■^ im » ^ee<rvM CMS *%.• * »
How to fill out your Census For.u
S## Ih# f'timd pul #a#mpt# in |h# yellow 101 |ruCti(K> (ju-d* th,; gv/idt w<l> h#'p «nth any ptQbl#m| VOU ma, h#vt.
H y3u M»#d mOf# help C»»I Ih# Cfniuj O'tir# Th# teleph«,n* p( ih# toca' nH<# •% t^own at mf pnttom ©Í |h# add'#)» by« on th# T'cni covet
U«# a h»ach p#ncii ro th# qveil on; 8'acK pfnet «i Cflte» 10 ui* than balipcnt o' 0»h#f p#r»i
lilt r"í.lf»S 0 COmplft#lY Ut th!) # W^fn »ou wM« >n #n |niA#' pnnj or wit# Cf i''v
Mak* >ü<í lhai «niivt'i a«# p-ovc^d 'c #v#ry»^' h#r«
S#* p^OP 4 of |ht yV'f* it I -r 'r"i#f c i f > rr . • i * # i i ' > -■«. -. - * "i i • i r% i .» to g -r , i »M ' .,• • ■<» <i • ""
At#wt# II r r "iiiioni ... '*■ I -h» •••< |h#n ir^tii^Q w.jh payf* 6 «>' O f I'd a o*-f ^ * tOi # ACh »f' n ,A |hf t.r,,,' » | • i
I h, n y-..i|# your n¿ ■ tt<fl anil * .,
Mill tack Ih i to"" on T^r$jav <♦ I o< J, Soo^ a*(K'Y*aid ai *on c#a U\* Of * "closed • no »!*n>p ,j i.p*(],d Pt:IN S'A I by a'-'iwt'-'-B DvcH^ ' '
Question 1
L*ft m Ou#|t on I •ÍAm.Jy l-v»oq h#f# n^hirt.ng bat) f 1 *;■>" ^ IM
' hvny hf,# * \ rw hfv«ti^»i U^r»<j h**» "Othff pr«soni KmIq h^p ■Co*l+q* Slvdfn*1 who 114V 'V am^vl.nq fo"#9*
if p^T»MS K-f ^5*w*^i# •fY'iCXH Wlto UK»>V try, N»'f tout #F# t^mpOfA"»y *w#y
i,nf i hiki en m tK%if<lrfV} y rx*> i^ ro^-j«
w^lh t hf rny plyWWf tM wM M#V h^f pf th# wrf+ wfb*# wini-nq
1 Whit l« <h« na*w# of •#rK p«r*pM who w#« Ihrtaf h#' & Ta#id»y April V O' who *» $(#y »ng or nt liit|) h#i • l,<Jr o'- *- t. ■ J
Do Nut litt m Qu#«tion 1 .Any peiw> #w#y I'om tn Ih# A/med fo<C« ,Ar*f c/j'trg» <ti»d<f>* w^O iTay^ i<xnewh#**« #lt# w+>ilt *:lemd'nq ro»^ne
. Any p*» vnn w+¥) u$u*Hy 5t»y* f'j# moil r' the wyH w^i'e WO*Iii«*Q Ihfi#
-Any c*»S'»n aw^y I»oti bt't *> A" mstriunon Such #t » t^orr* f(v thf surd CM mroia) N]S|i lat
.Any [VMsnn ji.T/tng o» vl«inp h#,c who ha» 4 v»u horn# lfo1« M #v€'yCinr hfi* >\ tliyng O^ly TpmfWV^nly j J I . ¿ ¿ uSiMt hprn# pif4Sf fn4»k !»«s bo» Q Ihmn pifas# ■ |nw |f^* qi!#il on$ on pvjc* 7 fhiouqh S ry»fy yd
. eniPf |hf A'VfcrM o< you' home on page ?0
h*iM centMue
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
140
indicated his intention not to issue a decision on whether
to adjust until late 1980, after the field enumeration was completed and when there might be some indication of
the quality of the census and, thus, of the need to adjust.
On 11 December 1980 the director announced that the
bureau would not adjust 1980 census population totals
at that time unless 'directed to do so by the courts. T^iis
decision was based chiefly on two factors: (1) the quality
of the 1980 census (the census counted some 226.5 mil- lion persons compared with the official estimate for I
April 1980 of 221.7 million), and (2) the absence of any
accurate measure of the number and distribution of ille-
gal aliens in the country. It was announced, however,
t^at the bureau would continue its research and evalua-
tion aimed at developing statistically defensible mea-
sures of undercount, and that if measures could be
developed that would clearly improve the population
estimates made between census years, future estimates
would be adjusted.
Several weeks prior to the director's announcement, a
U.S. district court, acting on a suit filed by the city of Detroit, ordered the bureau to adjust census counts, to
submit its plans for doing so, and to delay the issuance
of counts until they had been adjusted. On 29 December
1980, a similar order was issued by the district court in a
case filed on behalf of the city and state of New York.
These orders had the potential of delaying the issuance
of census data and preventing the bureau from meeting
the legal deadline for delivery of apportionment totals
to the president. At the last moment, however, the
Supreme Court issued stays to the lower court orders
allowing the bureau to deliver counts to the president on
31 December 1980.
In June 1981, an appellate court struck down the dis-
trict court order in the Detroit case on the grounds that
the city lacked the standing to sue. The Supreme Court
declined to review the appellate ruling in late February
1982, thus making that ruling final. In the New York
adjustment case, the appellate court had ruled in June
1981 to send the case back to the district court because
the original ruling (in New York's favor) had not protec-
ted the interests of other states. In early March 1982
the Supreme Court denied New York's request for a
review of the appellate decision. In a separate case, the
Supreme Court ruled in February 1982 that census
address registers arc confidential, denying plaintiffs
access to them.
Costs
The 1980 census cost over SI billion, nearly five times
that of the 1970 census (S221.6 million). Several factors
contributed to the large increase for 1980: inflation,
additional coverage improvement programs, other
improvements m the field enumeration structure that
indirectly improved coverage, new data needs, enhance-
ments in the geographic and processing operations, pop-
ulation growth, and an increase in the number of house-
holds.
A review of the yearly obligations of decennial census
funds, presented in table 19.3, shows that 60 percent of
the funds were obligated in fiscal year 1980.
Table 19.3 Annual Expenses for 1980 U.S. Census (in
thousand US$)
Decennial Census Obligations
1974 1975 1976 TQ 1977 1978 1979 1980 1981 1982 1983 1984
Total
642 2,204 5,180 2,568
17,073 29,090
132,472 664,642 135,183 50.606 41,289 18,887
1,099,836
NOTES: Figures through 1982 arc actual obligations, while those for 1983 and 1984 are estimated obligations. Fiscal years 1974-76 began in July of the previous year (e.g., fiscal year 1974 was 1 July 1973 to 30 June 1974); fiscal years 1977-84 began in October of the previous year (e.g., fiscal year 1977 was 1 October 1976 to 30 September 1977). The transitional quarter (TQ) is the period between the old and new fiscal- year systems (1 July-30 September 1976).
UNITED STATES
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
141
50 rN( » Juij '■ows ;«■ J VCA tEASS A \S*Efi QfiST.'OMS f<Jft YOUQ HOUSIHOID
H1-H12
l/NFJOSTAT: < SAMi'T^QUHSi'nnNA"»;:.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
142
--* ai se «íw.'íi tm wa/'jw, ol*í !*.'*•$ cw j
NOTE (;¡.'.-stion$ 1 • 7 and H1-H12 asked c¡ a¡] households. Questions 8-- 33 arvi H13-H32 asked oí sample huusenoids only.
UNI hi) STATES • SAMPLE QUESTION! AIRE
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
»#**#
II In «#Mt fU*a *r Mt tWf WWW
^ii*< m#* kM O» «w |M #* hmm #f • M M*M» )MW
*/ i*w t %f*#" tmi^rf, W Am»f* #**, Cwpw, a#a It <* #4 WHi' mm >*■ te * lai—fi -
» H Ma m<m» ■ wlwiM HWnm rf M* vww tlMna'
**t • «Hnfcifd M* (*|«n
n >»" atauail W M
fe ***«#*
1VÍ* IM0! ItMtolH»; lltWttttt MN| iwitiw; ' »HM i*
IJ« Dwi M pmm* WNk • *«f»i|i «*«r *#,
i •<» ##*&* (m#Mk - M* » U
k ****** ******
6ñ6,\ 6#"AT #Ñ) « M#* #*# #M Mi »»XW i***t fifijh'
14 WHH ta Mi ^mri ##«CMWy? *#*###»#
1'«^ * ■« <■!, Ww, fwmmum. ÍP»H Lf4 t*r*w»e KamawAm r* ;
I Si D**Ma h* in Mi Nw #w y#a f *t r«<i^ V A*mm4 f#rm te 4^» p#
®arn Ky4 Ifí W - tmm «* "•*< Tu m#t hnw - fA# »
(A*rl I 11«)» <l) tla*a
Pvtrt* ■*». — C*#m. Me
ítl CmtMf Oí («h
T**f *#T 1«) kw*M #* lnc<
tn
ara*aia« cw««ri M«
II Vkm tM Mi pnm« bmf ' > Sww N>vi IM4 -
** f* #• «>KM f t tf mm " Imi^iINtvMa- ® f— » »»if ipi i? ,*/**#*
• ta meM# Ma» I» IN* tan 1 **« U *#
o ?m « #w%h% •*»)■* nr tmtmm
'j VM WWm, o *» ' i *•* part #m#
It* * M| ww* 9 W#*## ## #pM* M#% M0lp| wwk» * #* HffiW >wt« W ## UiÉMl l«W # <■■*« ■* ■» m*r.
O M* — n »f fe ••» w#**n*to| mwhi Mu —
ftfm f**fc %» mí*
' (Aw^m tm* **,* tfri) '- r*brw*ir IHi-X#T l*M ', H»w ter*d f/*» ffW-^wy mi/ O »qr^ »*. W f! ■ ' I l tpáé f^ér tflTÍ
# +W»+ * l*r* !9IT.*,lwm*tt tttt) r> toif (#**» #m# I* I
• Vrf**» W w% i
km# a p^te^. *m *M. <m ###, *### ha* IM»«4 taf • «t
^ » a a i*— co «a ■<■>*» g o f*>wNinaw4»|M|^f O 0
té o o 2t frAUfwuiM - Mm | ( H t I
Nm liWai Km *M W f, O O O O O O kW m# e*vm#m# afewKN» 0# **r (##—f *•» ? # $ R II If
O O O O O O w t*a#a* i** k# *¿*0 n 4 f| #F #4 Am m
fVci f- % >>"» a*W f«« M»#» mé fmt
tá IW
t # *rW **w* Am* - Otf M # *>W bativM W #*» ####% #1 #* N
ANSwn rnrsf oursnom po* M" W*#a mw ti> »#$ «# #*, #me
o ?w - o *+ -
#- w#» ## ## vene* #w» W w# ###NWy
V) M htWwm 04 panw ###% *# i*aw*t»*nrti<^> W >*«4 »w« hiw i teatro w»i< hi» t*m
mwnW
% m*m #* C#v W#n. 'ii*a. I*iwif*i. mk
k #w «**c« «# wt '#"## «i ### «*|. «»##
ww, rn, >H| #a MW fW$#n
' U« «##* k## km# M # *w#* t«*« W# PMM a» #w Hm heme ta #*» |*m# **,p
&. *## ## Ma pan» mhmIII *ei «• MHi km#*#»»
O O ; tMKit G í"urt O MoMrr;* 0 V#m O fccyíl O iva m MMttf O w»a.«4 <r+f
*e*ned # heme ', m > 0**v — -
"n*. w K» » X*. f» i» >•«
r> (> <5 I 1 I
' > (' O I I !
¿f WUI uil Wf 1«
' t I I I '
S 5
o o o n r> 1*1 1 t I
*13 ■> > 'j ■, ) t 4 ', f, '» ' > r i > t
I '
" ■ n «• o o n ii. i\ o o ti! 1 -v ' i t i
r ( r t y , y ^ i •>';')
T, ^ G f. r ?
flTtíCW f ON PAGt 2
O Ck** aw* - f&» » f# >-»* ■*•••* w»t n W»a »-mm$ («trna^^w^
4 Nai M«f M"#*a Mi ####" « W «w% #*#«#(» frvck. # a#m W* —
^ } B • I ') J ■ 5 »•«
^iW «aa##m fw » /f n Im Ma pm*m» lampa H% ltn«< m m »## fr«*n * **
wMMMtajP**' O '« •" «•- 'm » iatiw. la nau *'> <Mw» Mw *
NIHNM ***** w*m ladina >•> #*#& fH* IMI « «#*» ■ r '« h» - i** *;' k C#wW Mi pmi *+*• i*»#, i
Pf»í »M* # /*• >"W •* I
Ta* c@vW Kw «a*#m a #& IF «U Ma paiii Im* «art m*n M» * l*« H#r#
1MB l'l 11*3* !•?< ^#7$ imtalt?» 1«M
5}%' M - M •» m##* p*4
Orm> *» ikvfy 0+ p»11 ^ ***f i** « W*^V T K»m*n ##*#
« laMiMNainMiiil'ff'n^aMA^kAf
k W*i* hWW a# taNnin «r Mw* I *a* »►"•'
y. wv , •»» » <0'«lf(K*< »
I K Mi wwimi»; — (rm arm t»i*i -i W ants'artunni ^ '• ftvtarf baria
a nada ^ Orna» - '« *$ tVrt »i| iMIi
I «IM IM «f aul »h Mi pMM Mén|'
ff'm i *^.*1 ' ^ *
& Ma i not icurtj ■< hn4w» #»
iW¿ M *»• Ma iinm - (*m •—» (w*)
tuyhijM a* evaa» ro^wr Kr»«m* «■
laf ^*r. i a IMI ** } la* #» p-#« # *4 iwy D> K Nr « or t*(" —
P*« M"«aii mal *<w»rtrt
r k« >wan «r *
CfNUrt tm
II»
y y * i i
M k I C
o e f
0 M J
■ l w
W P Q
# 1 ?
u "» W
« T 7
IU iMt m» <1*ti «4 Mi —■—n #v» m*" Mf*. «• e ##W *& m * a WW an » Mm'
fe Mm *pi wn&i Mi piran* M tn l»r CM« pnW fW n* Wan. ^»1 i #m
t #**# A# ###Ai hi<* n» i»r» i^# tmm* m*
# 0* #* a^« » 1*1^ (# N## ■ »#a Mi hww Wnkim# *•» ##% ^ «# kw" Wm» » n
Clfcl' l V1I o*«« M% )M "ll<
t • ;
U kww un - m>00mm AM mam'#
H nmt p«M A >'—■" ***
imMMil WWnl ^
1K >M • f*it Mn rt"
fe O*# p»ne*t* «w w^n#
)h »
O «- I
$ lat* % m linmtel
(««•■i r»*h kté te r (•#0O # mmm
IH II
J| *km ##, Ma
*» ^>W« km * >» r*rt PH« irV w» *1 txiO»» *r ^|»| ) tu #«*»
NOTE: Questions 11-33 repeat on pages 8-19 of the questionnaire for other persons in the household.
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
H tT) a C1
Í r.O
C/i >
r m .O c m in H
C
?3 m
P#*« 4 !"ii •»
MSO ANSWf* THfSf OurSTtONS xh WU jw'omi P*W fcu*hnf* W(f«vMMiw 4Mt fH **«•*
A i'- NortT» u h*l|i * tvw m«wW OP'K'^O t-wn *** MM# WwAA 4 <m« NV.1I Afw f |d to *m# # mM htiMM A bu»k>i>g W ? I»M^ A (Of ) a» 4 A 't« % to ♦ I■»**•<HH A tn«"iW( te» 10 1u 11 A k,'W tw JO «9 At A tu» 50 9 "*»» '*"*m
H^4« H## m#A, 11«««« ffWvn} «r« tn í <w-ii m*tm #» m $ nm¡
I k» J - i*# NMM «hit
ff* hutttf»*' »' »/ %# #my A
7 to 17 I) or A*t man*
% It **f* A n«*IH« *» AH
» kWdk# - On p city fw AuCKvtv»" kgt gr on I plf* f O" • tJ«;# t* I to Í AtFM* !>• j o< 10 > "*w* Af'A»'
ton l**rt 1 acir* - )»# <• «'I
Mil* IW «a «ttf MM ## h#wm hn»mf 6*4 »w p«rt Cm bOfHw) <#* «r L# l>actric*i r <«• 0rf kfurn **
CAM * COk* OMf K»M
• • N» kiM tnatf
k Wttitti KmI M tr**4 m*M Im »#W G»t hmm pv*$
w*v<n« the wghhrnrfwaal > QM homw i*r* « IP
fu* 9* MrcMm #* 0*»#f km# N» k# v*N
« Vfctt* K*M n w*w -mm H* c«o*tnff ImM from i^virfnuntf P»W*
Mrwn lh* r*^N**»*od CM brWWd M«fc. ■>
fe tWU !*•». Iff <,l i WM Wf Cr«*| f W»n Vua HM • HPswM ## —
Itnl^r'ityiivN'v^l If V* *#!'?* 1*0«. U*9 g (Mkim
I Mri *#** >Mm #ve**h
li ano +\3,t U *tt># m«n
| &• yaw kwft - A pvWf /, (, jtpmtawiii, M(J # »r>.r Ar *«IK Ar, m*#'
M)/ I. * . jtfMiAf <0^n*Ctvd ko • | ton**(**d la pwWf w#*
í)o ¿orrtKIM t0 MClU l*AA (? (tlUTQ' No wfc* an+r
Hjl *n^M •'• lh« CM<| «# ^tw## 1*4 K**ü fw f*ur 4Mn« # tlftlnclH I 00 0* hxhidad f <*<* » <»
[***<*, noI ual k Cm
t kUM >A iwrt m m CM wwd
» WkdW M rvn flp m# rh«rp»
4 ON. CM(. ket-wne. w*W. #W » 00 0« IncUM ^ t0 no cK*rp|
Wl <%* Hj] 0# y*w ham kikh#n fA%W#**f XI<*M
««(■Miwi'tMMatNr.i'Wfi ■
HI | Kf#w< «ftw «»■ ttwt ^wiWImg wifIMM' #WA -*»-> f*# , " rtni ■ an!i'-M (•# «M ***** rt MMW t» J» iaHwlM
J*F< or I9*n I1IS» r. 'o io i»;*
i*60* IHÍ !■>»*> 19M
i»*0 w IM1 I9>t ar
"1? WK*.> d»5 A* ##'*** kit#d h r-.ii heve* (Ai Apéf*wwi<)f
l*my 19W i^Tiio nil <4/0 it.
I iiwm Ml
IWOWtV*» 1941 9 r#'U#
' mJ h*f*
4 How Tk*ny &***#; m# ## fov k#m P ÍMM i' —ni M »*—«*> /«r I***"# mw » aW #4* <*f ««h^ ^iifin
Mo b»»m^ ? bOT>oonn • * b*dform« 1 h»<>y»n J hrt>BtfA| 1 ) V mcrt
Hft Hn m*Af kwh'Mww éú r«u k*wf ■«* •*» A Wf Wl* ,"#* A#, «< 'MN « taAN ff- &WA*A m mt *m* #A»hi •i'i'r. *mf 0 NAw
Ho B» ftr^y fl h*N h4fN'*t»A 1 ha**wOQm I fWfvmqr*. |A<4 h#$ - ] «• trwjit c4r*p*H W^#or*$
MM 0# r*u Kt«« » m y#w# #*lm# «u*ftan7 '« n *•
.tJO Nc rei» i«>rt| *wAn*f# hfiWf If 'w I'l A#»* uW #*|
*>lr« " * »-o« »#t«f ^* tiff Ay * »'(M 4^1 M ^
fUm «ni rKitrH A«H htn) li*rr K ho# eufr-v '.iV bv.W in •!*!*< uMt
i Itm» m*M or f^f*» Mn. .,, i.^ or wfr# {H O* 0» Wr»Wm# •r. m m»Ui\*"'(• « v«N. Uj«rt.n( (M »1 *r fcoroX'^» 6 í«rnJ«íP> *« room h*4«*rt 0* »"W **C r»*r»f p*-ipm*.4
^ Hj7 &t r*M kw Ak C0«wirt^»wgT *•* A *•' i^ftd«KK»ni HriiW *n I nd>w^i» raw uM Tfi l m ***** *<!■■• . .: -ojm
Htt Me# m#m; A^imUH *r« k«« »| hMM MtiMbf MMim»,,, W y*u* k#w#ih##W7
Men* | ?A«waMii | 1 »i<«»mri»W • i «r *>#* #w*@p*#Wm
HJt Ho« fHA^r mm «r «rucki o< mm* ** CA#*cM> •» Ini #m k*9* A# hwne W» tua hf wwfféi'i ¥ fmm Aeimi^ilrfT
wm I <««»i # kucti 1 MM Or m#-k ' ) «r W| i«M AT tUT**
C(MUI U)(
Milk
/DA YOtt* HOU51HOL 0 FSm* m*r**w HK> Hj? tf fhm Im * *m fmmS/ Mmmm
fOM « ** h/y4»#. fArn gk4 # - • A wo4*A« ham# « *»*##» " # >>MM #rn 10 #r **# *cr«a • I tomdlj*' MWWfl M* • A *K»"+4 *#»* a cp'Mwiid AfUbkAXm**
m mWK a> ##*# #m I
# V ##* « «r "**« *»• #r * A mA>V«<rt*«ii«n Akf*M ANiJM'w'irmtl
MM WHaI ##r* #** f*w Ml«M ***** m (Ü| wn***! *W p**P
I 00 CM »9**
Mil WHm W #*# #mm*W r*mhM Im #md *ml*^ >hmiwi #* *mj
» 00 C* No»*
HM* 0* fU h+n* • m-rt»*®# áná «f M wnfid * pw(fc*«a # Mm*#
*r* 6##d #* ir\M •» W"W #tM '. TM, tVAKl k »ur(>i«w
k #» Mw ## ¢4 ##«##*#» r- v«# o m ». •#
■•WVWsV^
< M*« r*wt*» H rwwr «M# '*#V&#r Mf
1 00 0* M» • «#L"Af *»>■'« < '>»»*< - >■
Í p**< r*^ '*1^* ««•"•*** f»»»iw<rt M»K I# MJfd m'UA Hr*Ml w ná MlMl 1**4# #m »MJ #r**#r1;l
*% l»*M Mlw»A #" 0% *•« I*MI Mv«<Vr »
' .- NW'1 NWAI mm# Kii»^ wiwi #* £-) pin -*y*
M rwm
- (
^Att in 9 »
f0# ctw*ui uw (#*r . W \ .
O 11
0' :' 0' '
® I »
0;' ta
© 11
CO ■ H» . j
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986
PI*### Make Sura You Hav#
Filled This Form Completely
&
An$m»f§ m W, MJ. HI Nem# a# W#
Ml
W #"a*a at i »ii'W wNein #**# # ## #m# # m# h#** IÍ» liUmpX N pfMA ## » Cw»> ?M
MOT1 M you have lifted mor« than J p#f$on# m Ou«it»o« 1. pitase make sure thai you have Idled the form lor lh# l»rit 7 people. Then mad back thu form ACtfliul T#Wer *WI c#M 1o oblemtht informMto* 1<k rhe other peoo$#
CM ## k# te<fiH y#w h**#: * AnyM^red Ow#$1*OM 1 OA pap# 1 * An$w##ed 0u*tt<O*« 2 through 10 (or ##ch m<*o* you
fcfted #1 the loo of o##** 3 #*d 3 * Art«M«redOue>i»on9 hi ihfowghM37 onotQ» 3. * end5
* Filled a pair of peges lor each perion i>«ied oo page# 7 and 3 Thai »» page# 9 en<J 7 mhouW be 'or t*» P**$om »n CO*vmn 1. pa0tt 0 end 9 lor lh# Pfrton m coiv«>« 7. etc N*$# «0*»« «* »m*## I I» atMli mm* 1 T )} **•«9**«™ to'erf **"* I iM »rw^
w KriM *• ## »K# CCV« Wt •» ífl «c(Wi'o« I IH*4gi' (0'"Q ^ 0« »»•«"•{! r**io* T@ mt Niwif io c^*ct nu ie «»«« o**### w C#1#m V«M h#** |«4A #1 #•« n^CMtlT
Writ» h### tha nema of lha panon who M'ad tha *o,m lha data tha form wa> comotatad a"d IK» r>umt>»f on wh»ch tha paoiMa hi thu houiahold can ba cañad
Tl»a« fold ifca fotm th# may fi w*i ssrM to you Ma>l ,1 bach m lha anc<os*d arwaiooa lha addnn o< lha U S t»uui O^ma apDaci on iha troni cover o' ovps^o^m#,fa Plaaia b« luia that ba'o»» yousajitha envelope ih« add' a» •how* through lha *tr»do» No llamo i* raqwed
Thtnt fov wy mart
.U cn
INE
GI.
Mem
oria
: Ta
ller I
nter
amer
ican
o de
Eva
luac
ión
de C
enso
s de
Pob
laci
ón y
Viv
iend
a. 1
986