Overview of the 1995 Census of
Population and Housing

in Israel *

Benjamin Lasman

Census and Demography Department

Tel. 972-2-655 3204

FAX 972-2-655 3531

e-mail: maria@shum.cc.huji.ac.il

Israel's first census was held in November, 1948, six months after the
establishment of the State. This first census formed the base for the National
Population Register and the subsequent censuses held in 1961, 1972, and in 1983
made extensive use of information in that Register.

'Census Day' for the 1995 Census of Population and Housing was November 4,
1995. The census covered the population living within the borders of the State of
Israel, and the population living in Jewish settlements in the Administered Areas (the
West Bank and the Gaza Strip). The population to be enumerated was estimated to
include slightly more than 5.6 million persons in 1.6 million households.

I. Enumeration Procedures in the 1995 Census

The 1995 census was conducted using a delivery and collection self-enumeration
system, based on a drop-off and pick-up field procedure. This two-stage approach
aimed at achieving a high degree of coverage by means of a repeated complete and
systematic canvassing of the Enumeration Areas (EA's), carried out separately
during each of the two stages.

The drop-off stage:

The purpose of the drop-off operation was to prepare a complete list of all
buildings and all dwellings in each EA and to deliver questionnaires in the dwellings
to be completed by the residents. Beginning ten days before Census Day,
enumerators canvassed the EA's which they had been assigned to, distributed
questionnaires to every dwelling unit and asked the families to complete them. At this
stage they were guided by detailed maps, prepared by teams of GIS, separately for
each EA. The maps showed each structure, street, street name and code, and house
number. The enumerators were instructed to proceed in their canvass, strictly
according to a route which had been drawn on the map in advance by the
supervisor, in such a way that it passed through every structure to be covered.

*This workshop is dedicated to the integration made in the 1995 Census of
advanced

technologies. These technologies are not elaborated on in this paper. However, the
following papers present this issue in detail

Two basic questionnaires were distributed, a short form and a long form (see
section IV). The long form was delivered to every fifth household, according to
sampling digits which were specified for each EA, based on the serial order of
dwelling units encountered by the enumerator in his canvass. In addition to
distributing the questionnaires, the enumerator recorded in the Enumerator Report
Book (ERB) each structure and its address, as well as each dwelling unit, the
name of its occupant and a serial number of the dwelling, which was unique in each
EA. He also classified the dwellings in the EA (e.g. dwellings inhabited, vacant,
occupied for other purposes than habitation).The enumerator was instructed to
maximize coverage of households by leaving questionnaires at every possible
dwelling unit, unless he was told unambiguously that a particular structure or
dwelling was uninhabited. Thus, at the end of the drop-off stage, the enumerator was
to have delivered the appropriate questionnaire (short or long form) to each dwelling
unit, and to have prepared a complete record of the dwelling units, whether occupied
or not.

From the point of view of coverage, the advantage of the two-stage procedure was
that in the drop-off stage the enumerators were able to focus on coverage of
dwellings, without being bothered by contents of the questionnaires.

The pick-up stage:

Following Census Day the enumerator returned to collect the filled-in questionnaires.
If the household members had failed to complete the questionnaire prior to the
enumerator's return, he was instructed to attempt to complete it on the spot.

At this stage, each enumerator was provided with an auxiliary list of the persons
who, according to information in the National Population Register, lived at every
address in his EA. When the enumerator returned to a household to collect the
questionnaire he compared the names of household members entered in the
questionnaire with the names of persons who, according to the National Population
Register, lived at this address. When the enumerator located a household member on
the Population Register list, he removed from the list a self-adhesive label, on which
the person's name was printed, and stuck it on the questionnaire, adjacent to the
information which had been entered for that person. The auxiliary list served as an
important independent check of the coverage of persons in enumerated households
(dwellings). The use of the National Population Register in the census is described
in more detail in Section II.

The pick-up stage lasted five weeks, and enumerators returned at least three times
in order to obtain completed questionnaires from households whose members had not
been at home during previous visits. A final effort was made to obtain completed
questionnaires from households who had been visited three times during the pick-up
stage but from whom questionnaires had not been received. Enumerators returned
for three more visits to these households and tried to obtain, at least, information
on their size, either from the household members themselves or from neighbors.
This final stage (the 'clean-up stage') lasted up to one week, and with its conclusion
field work ended in all EA's.

The basic enumeration procedure was adapted also for enumerating special
populations whose living arrangements or geographical distribution made its
implementation difficult or unnecessary. These special populations included residents
living in institutions (such as: boarding schools, students dormitories, homes for the
aged, prisons, etc.), residents of kibbutzim (collective settlements) and households
living outside the municipal boundaries of localities (primarily Bedouin).

II The National Population Register and the
Census

The link between census procedures and the Population Register administered by the
Ministry of Interior is a unique feature of the Israeli census. The utilization by the
census of information in the Population Register has two main purposes:

1. Improving Coverage -

The use of the auxiliary list described in the previous section led to an improvement
of the coverage by, usually, correction of an oversight of a person (or persons)
by the household and his (their) inclusion in the questionnaire. In some cases, the
use of the list led to an inclusion of an entire household and even a whole building,
omitted during the canvass.

2. Facilitating Editing -

The pre-printed label attached by the enumerator to the questionnaire contains a
unique identification number (including a control digit). This number which is
subsequently captured with the data in the questionnaire, becomes part of the
person's computerized census record. The inclusion of the identification number in
the questionnaire makes possible the record linkage between the individual's
questionnaire record and the corresponding record from the Population Register.
This record linkage enables the use of the demographic information appearing in
the Population Register to supplement missing questionnaire data and resolve
response inconsistencies during automatic editing and imputation operations.

Additional uses of the Population Register include evaluation of coverage and
quality of response.

III Census Geography and Field Organization

The basic geographical unit of the census is the Enumeration Area, which contained
an average of 260 households. Prior to the census, detailed mapping of the
country's area was carried out and geographic EA's were defined. Each EA was
assigned to one enumerator. About 6,000 regular EA's, 2,300 institutional EA's and
280 EA's located outside localities were finally delineated. Sixteen census
administrative regions were created, and each was subdivided into local census
offices responsible for enumerating an area containing some 40,000 persons. Each
local office supervised 40-60 enumerators, divided into 8-10 crews. Each
supervisor was responsible for the enumeration of 6 EA's.

Most of the organizational, training and enumeration tasks were carried out in the
local offices, while the regional office staff provided support services, especially
with respect to the recruitment of nearly 10,000 temporary field workers, paying
them, and overseeing local office operations. National census headquarters were
established at the Central Bureau of Statistics offices in Jerusalem.

The network of EA's was also the base of the hierarchical division of localities
into geographical-statistical areas according to the principle that each larger unit
was composed of smaller units: 'statistical area' (in towns having at least 10,000
inhabitants), 'sub-quarter' (in towns having at least 40,000 inhabitants), 'quarter' (in
towns having at least 100,000 inhabitants). In most cases comparability was
preserved between statistical area boundaries as they were defined for the 1995
Census and the corresponding boundaries defined for the 1983 Census.

IV Questionnaire Content and Design

Questionnaire content was determined in the course of a program of consultations
and testing, involving experts in various subject areas, recruited from the staff of
the Central Bureau of Statistics, and representatives of government agencies, public
bodies, universities and research institutes, organized under the auspices of the
Public Advisory Council on Statistics. A balance had to be kept among often
competing demands: meeting new data needs, insuring comparability with prior
censuses, adherence to United Nations recommendations, and preventing undue
respondent burden.

The short form contained five demographic questions addressed to each household
member. The long form, delivered to a sample of one-fifth of the households,
contained an additional 25 questions addressed to household members aged 15 or
over. Topics covered by these questions included education, geographical mobility,
marriage and fertility, labor force characteristics, journey to work, and income. The
long form also included questions about housing conditions and possession of
durable goods.

The two basic questionnaire forms, which were designed for the enumeration of the
vast majority of the population, were also adapted for enumerating the population
living in institutions. Special attention was devoted to the graphic design of the
questionnaire, in order to make it clear and attractive to the population and thus
maximize the desired self-enumeration, and in order to meet the special requirements
of the optical data capture. Official census forms were printed in Hebrew and in
Arabic; translations of the questionnaires into other languages were made available
to the enumerators to assist them in the field work.