Publications

Progress Report on 1971 Population Census of Indonesia

Presented at the Workshop on Population Census Tabulation

East - West Center, Honolulu, February 12-17, 1973 

Sam Suharto and M. Abdulmadjid

Central Bureau Of Statistics

Jakarta -- Indonesia

TABLE OF CONTENTS

INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .1

PREPARATORY WORK. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

HOUSEHOLD LISTING AND SAMPLE SELECTION . . . . . . . . . . . . . . . . . . . . . . . . . . . 12

CENSUS ENUMERATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .16

DATA PROCESSING AND TABULATION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24

EVALUATION AND ANALYSIS OF THE RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . . . 39

 

1971 POPULATION CENSUS OF INDONESIA PROGRESS REPORT

INTRODUCTION

Indonesia, a country in South East Asia consisting of 13,000 islands of which only about 3,000 are inhabited, is the fifth largest country in the world in terms of the size of the population; the four countries whose populations are larger than Indonesia are China (the People Republic of), India, the United States, and the Soviet Union. Indonesia obtained her independence on August 17, 1945, after being colonized by the Dutch since 1595 and by the Japanese since 1941.

The present Government of Indonesia is headed by President Suharto who replaced President Sukarno in 1966. Sukarno had been the President of Indonesia since Indonesia became an independent nation in 1945. The country is divided into 26 provinces and each province consists of several regencies and municipalities; each regency is divided into sub-districts and each sub-district consists of several villages, and in many areas the villages are divided into smaller sub-villages. In the entire country there are 287 regencies and municipalities, 3,171 sub-districts, and about 45,272 villages.

The Government launched the first five year development plan in 1969, realizing that one of the primary requirements for planning and carrying out the economic and social development for administrative activities and scientific research is a set of precise and detailed data on the size, distribution and composition of the population. The undertaking of the Population Census was, therefore, considered of utmost importance and was included in the framework of the first five year development plan.

The importance of Census data was also recognized by the Dutch colonial government in Indonesia which conducted several censuses, although only those of 1920 and 1930 were considered proper censuses in terms of the objective and the method used.

The Indonesian Government conducted the first Population Census in 1961 as part of the 1960 World Population Census program of the United Nations. Due to various reasons, partly economic and political, not all of the tabulations planned in the 1961 Census were actually produced. However, this Census was a very valuable guide for the planners of the 1971 Census and many aspects were improved for the 1971 Population Census. This improvement was eminent in the data processing stage as a result of the advance development in computer technology.

The planning of the 1971 Census started as early as July 1969 since it was realized that a population census is one of the most complicated statistical operations, consisting of various steps which must be carefully planned in advance so that proper control of the continuous sequence of activities can be implemented. A master plan was drawn up for the whole operation of the census covering the period 1969 through 1974. A population Census as defined by the United Nations is "the total process of collecting, compiling, evaluating, analyzing and publishing demographic, economic and social data pertaining at a specific time, to all persons in a country " (Principles and recommendations for the 1970 Population Censuses, Series M - No. 44 United Nations.)

The 1971 Population Census operations including the above mentioned activities may be categorized into five stages as follows:

  • Preparatory work
  • Household listing and sample selection
  • Census enumeration
  • Data processing and tabulation
  • Evaluation and analysis of the results.

These five stages are interlinking since the outcome of one stage becomes the input of the next stage.

Top of the Page

PREPARATORY WORK

After the Government approved the overall census budget in mid 1969, the preparatory work of the Population Census began. This included formation of the necessary working committees, investigation of the existing legislation pertaining to the conducting of a Census, strengthening the field organization, formulation of Census questionnaires, preparation of instruction manuals for the enumerators, conducting the pilot Censuses, planning for the data processing, printing the Census forms, etc. These are briefly described as follows:

Committees

Two committees were formed by the Director General of the Central of Statistics. One committee was the Interdepartmental Working Committee which had the task of finalizing the plans involving methodology, topics to be included in the Census, the concept and definitions relating to characteristics in the Census, etc. This committee was vital since this was the forum of subject matter specialists who would explore all the possibilities so that the Census would produce the maximum information for cost expended within the limitation of number of topics which could be included in a Census. This committee was vital since this was the forum of subject matter specialists who would explore all the possibilities so that the Census would produce the maximum information for cost expended within the limitation of number of topics which could be included in a Census. This committee was composed of senior officers from the various departments which would in the future be the consumers of the census data. Another important function of this committee was brought about due to the opinion of various government departments that the Census should include as many topics as possible of interest to them, while, in fact, only a very limited number of topics could be included in the Census. Priorities were then established and approved by this committee.

The second committee formed was the 1971 Population Census Steering Committee, which was composed of senior statisticians within the Central Bureau of Statistics and which was responsible for the overall operation of the Census, such as the organization, administration, sampling design, questionnaire design, instruction manuals, system of enumeration, data preparation, data processing and tabulations, publication, analysis of results, etc. Each member of the Steering Committee was responsible for the successful completion of certain operations and a PEFT network of about 1,000 activities was employed in order to manage and control the timeliness of the Census project. The Steering Committee was chaired by the Director General of the Central Bureau of Statistics: this committee, which met at least once a month to review all the activities during the preceding month, could formulate new policies or apply new strategies to overcome the possibility of late or almost late activities after studying the cause of the delay. In checking the overall activities the committee was assisted by a computer printout of the PEFT analysis which was prepared monthly just prior to the committee meeting. This technique was found to be very helpful.

Legislation

It is customary to have valid laws and regulations providing for the conducting of a Population Census, which is of such magnitude and importance. Indonesia already had a Census Law, i.e. Law no. 6, 1960, and also a Statistics Law, i.e. Law no. 7, 1960, which provided for a Population Census to be conducted; however, as it was mentioned in the Census Law that every time the Census is to be conducted it must be supported by separate government regulations which are concerned with the execution, organization and procedures of the Census, several other orders and regulations were enacted for the 1971 Population Census. The most important ones were:

  • Presidential Order no. 2, 1970, directed to the Minister of Home Affairs and to the Director General of the Central Bureau of Statistics, to conduct the 1971 Population Census

  • Government Regulation no. 29, 1970, concerning the execution of the 1971 Population Census.

Questionnaire Design

Within the Steering Committee a sub committee was formed which was responsible for the preparation of the draft questionnaires. The sub committee worked very closely with the Interdepartmental Committee and was also responsible for the preparation of the instruction manuals for the enumerators.

Three pilot censuses were conducted in January, June and September 1970. The purpose of the first pilot census was to test the type of form to be used in the Census: the conclusion was that a separate form would be used for each individual in the household in addition to one household form containing questions asking general information on the housing and household characteristics. The second pilot census was aimed at testing the content of the questionnaires, the duration of the training period needed for the enumerators, the optimal work load of the enumerators, the ability of the enumerators, the time required for interviewing a household, etc. The third pilot was conducted in order to finalize the questionnaire and to determine the qualifications needed by the enumerators in order to be able to successfully conduct the enumeration.

Urban and Rural Classification

As in the 1961 Population Census, the determination of urban and rural areas would be done at the village level. A very simple form was sent to all villages, classified as rural based on the 1961 Census. A "yes" or "no" answer was required from these four question;

  1. Do most people work in the agricultural sector?

  2. Is there a hospital or clinic in the village?

  3. Is there a high school or higher education building in the village?

  4. Is there any electricity in the village?

A village is considered urban if the answer to question 1 is no and the answer to the other questions is yes, with the exception that all villages in the Jakarta area were considered as urban villages.

Most of the villages returned the forms to the central office, however, there were many which did not. When these forms were not returned, the local chief of the Census office was advised to follow the above scheme in determining whether a village should be considered as urban. By using the above-mentioned criteria, 2,668 villages were classified as urban and 42,604 were classified as rural. An evaluation of the above-mentioned criteria has not yet been carried out, but it has been suggested that a study be conducted to determine the best criteria for classifying an area as urban or rural.

Field Organization

When all of the planning, which was done in the Central office in Jakarta, was more or less final, a briefing was conducted of all the Provincial Census Directors which covered all the field operational aspects of the Census. A rigid time scheduling was given to them and one of the first duties of the Provincial Census Directors was to conduct the household listing and house numbering and at the same time to form the Census blocks in their provinces. This operation would be integrated with the voter registration in preparation of the 1971 General Election. The Provincial Census Directors then conducted briefings of the Regencies Census Directors in their respective provinces. The Central Bureau of Statistics also has one census staff in each sub-district who would become the lowest-level organizer of the census operations in the sub-district.

In addition to strengthening the field apparatus, the Provincial Census Directors were responsible for the recruitment of the enumerators for both the complete and the sample enumeration, the setting up of the training centers in various places in the province, and distribution and recollection of the Census documents, publicity, the financial administration, etc.

Data Processing

There were two systems of enumeration in the Census. One was the complete enumeration using a very simple form and the other was the sample enumeration for more detailed information. The processing of the complete enumeration was carried out manually in stages starting with the lowest administrative unit and going through the province level and then finally the overall total was computed in the central office. Various checks were designed to minimize the calculation errors. The population and the household totals from this processing would be used as the base population in blowing up the characteristics in the sample.

The result of the sample enumeration is being processed by an electronic computer. A feasibility study was conducted in 1969 with the help of a J.N. regional advisor to determine the various alternatives for processing the census data. It was decided at that time that a new medium-size computer should be procured in the very early stage and capabilities in the field of systems and programming should be developed within the Central Bureau of Statistics. The cost of the computer was incorporated into the Census budget. Requests for proposals were sent out to various computer manufacturers and presentations were conducted in December 1969 by all the invited manufacturers. An ICL 1903A was finally ordered and installed in July 1971. For the software and personnel development in the data processing field, the Bureau contracted Computer Systems Advisers (CSA) of Singapore, with whose help the 1971 Population Census system was designed and the computer programs were written before the census taking began. The computer hardware configuration that was chosen to process the 1971 Indonesian Population Census was as follows:

  • ICL 1903A processor with 32K words of core store (1 word = 24 bits)

  • 6 magnetic tape transports of 60 Kc/s

  • 2 exchangeable disc store transports of 208 Kc/s

  • 2 Universal Document Transports capable of reading 150 documents per minute

  • 1 card reader, reading 1600 cards per minute

  • 1 line printer with 160 print positions and operating at 1350 lines per minute

  • 1 card punch capable of punching 300 cards per minute

ICL has produced a software system for the analysis of massive files, such as census, known as FILAN, an acronym for File Analysis. FILAN is a system for checking, editing, and processing and tabulating a large file of data. FILAN was the basic system used for processing the Census data.

Prior to the computer processing the documents went through a data preparation stage which started with the arrival of the census documents from the various parts of the country and ended when the documents were fed into the computer. Included in this stage would be the manual coding of the occupation and industry characteristics of each individual.

A team consisting of computer system analysts and demographers was formed to design the editing and imputation procedures which would be incorporated, after necessary modifications, in the computer programs to clean the data.

All of the data processing planning was done by a subcommittee formed within the Population Census Steering Committee.

Preparation Review by Advisors

During this preparatory stage the Central Bureau of Statistics invited several United Nations advisors for short visits to Indonesia to review the Census preparation. Discussions were held between these advisors and the Steering Committee members on various aspects of the Census operation. Those who visited Jakarta were Dr. K. Horstmann and Miss K. Jubb, both of whom were regional advisors on population statistics of UN ECAFE, Bangkok, Dr. V.O. Kannisto, the interregional advisor on population statistics of UN Headquarters, New York, and Dr. D.E. Peterson, a regional advisor on data processing of UN ECAFE, Bangkok. Many useful recommendations made by these advisors were incorporated in the Census planning and operation.

 Top of the Page

HOUSEHOLD LISTING AND THE SAMPLE SELECTION

To ensure the coverage in a given area and to avoid omissions and overlapping it was necessary to construct a more or less uniform territorial unit called a Census enumeration block (EB ). Every village throughout the country was divided in EB's such that each EB was composed of:

  1. 60 to 125 households in the rural areas of Java and Madura.

  2. 30 to 70 households in the rest of the country (that is the urban areas of Java-Madura and all areas in the outer islands).

The main reason for this distinction was that the characteristics of the population in rural Java-Madura were more homogeneous so that it would be advantageous to have larger EB's

In the formation of EB's it was recommended to follow the boundaries of the existing smallest administrative units such as the sub-village and also the natural boundaries such as rivers, roads, etc. The Census block forming was conducted in June-July 1970 and was integrated with the preparation of the 1971 General Election. The size of the EB coincided with the requirement for the size of the electoral area.

Household Listing

The household listing was conducted for the purpose of block forming and for registering voters for the general election. In this listing two types of forms were used: one form used to collect information needed to form the EB's contained questions such as type of building in the area, type of household (institutional or private), number of household members (Indonesian or foreign), and whether any member of the household had an agriculture holding. The other form contained questions on the age, sex and marital status of the household members for the purpose of constructing the list of eligible voters.

Mapping

Another important activity conducted in this stage was the construction of EB maps (not scaled). Those maps included all the important land marks in the area, the lay out of the buildings and the corresponding building numbers. These maps would become an important guide for the enumerators during the census enumeration.

In addition to the EB maps, the Census officers were also requested to prepare the village map with all the EB's boundaries within the village as well as maps of the sub-districts and of the regencies or municipalities.

Sample Selection

The main purpose of the household listing and block forming was to construct the sampling frame. Once all EB's were constructed and numbered, a list was made for each regency and municipality in serpentine order by sub-districts and by village. A sample of EB's from each regency or municipality was then selected using systematic random sampling. All households within the selected block would be completely enumerated by using the sample questionnaires.

The sample size of each regency or municipality was determined as follows:

Stratum Total Population Sample Size
Regency 200,000 or more 2%
  100,000 - 199,999 5%
  less than 100,000 10%
Municipality 100,000 or more 5%
  50,000 - 99,000 10%
  less than 50,000 20%

 

The distribution of sample blocks by province is shown in Table 1. From this table it can be seen that the number of sample blocks was 13,793 from the total of 361,843 EB's. Therefore, the overall sampling fraction is 3.81%.

The household listing and the EB formation was also useful in estimating the requirements of materials needed for the Census enumeration, such as the number of questionnaires to be printed, the distribution of questionnaires for each region, the number of enumerators and supervisors needed, location of training sites, the budget preparation, etc.

The sampling frame constructed will also be very useful in the future as the frame of other sample surveys. Since only a limited number of topics could be included in the census form but detailed data on social and economic characteristics, which are not appropriate for collection in a full scale census, are needed, a continuous program of intercensal sample surveys is often conducted in Indonesia which needs a substantial sampling frame.

 

TABLE 1: Number of Census Enumeration Block and Sample Blocks by Province

Province Total EB Sample EB Percent
1. Aceh 8,670 302 3.48
2. North Sumatra 23,997 1,006 4.19
3. West Sumatra 10,980 402 3.66
4. Riau 5,566 202 3.63
5. Jambi 3,823 249 6.51
6. South Sumatra 11,560 381 3.30
7.Bengkulu 1,998 119 5.96
8. Lampung 9,811 291 2.97
9. Jakarta 20,043 1,008 5.03
10. West Java 54,341 1,774 3.26
11. Central Java 54,491 1,893 3.47
12. Jobjakarta 6,838 246 3.60
13. East Java 67,112 2,392 3.56
14. Bali 8,035 319 3.97
15. West Nusa Tenggara 9,702 316 3.26
16. East Nusa Tenggara 8,140 379 4.66
17. West Kalimantan 7,660 182 2.33
18. Central Kalimantan 2,671 174 6.51
19. South Kalimantan 7,485 382 5.10
20. East Kalimantan 2,320 102 4.40
21.North Sulawesi 5,505 211 3.83
22.Central Sulawesi 3,271 113 3.45
23. South Sulawesi 18,390 969 5.27
24.South East Sulawesi 2,532 127 5.02
25. Maluku 3,755 191 5.09
26. West Irian 3,147 63 2.00
Total 361,843 13,793 3.81

 Top of the Page

CENSUS ENUMERATION

Enumeration

The complete enumeration was conducted from September 6 through September 24, 1971, using a simple one-page form containing only questions concerning age (grouped as 0-4, 5-14, 15-24, 25 and over), sex, and citizenship. There were 160,120 enumerators and 33,148 supervisors for a total of 193,268 persons who are recruited in advance for the complete enumeration. The recruitment was done by the sub-district census staff according to the directions given by the central office. These directions included the number of enumerators and supervisors for each village in his sub-district and the qualifications of the enumerators; it was specified that these enumerators should come from the villages they would enumerate. The training of the enumerators lasted only one day, since the questionnaire was very simple, and this training included discussions on the interview method, some concepts and definitions, instructions on filling in the questionnaire and also the field try out.

Indonesia used a combination of the 'de jure' and 'de facto' methods in the census enumeration. For those persons who lived in a permanent place or house the 'de jure' method was used for enumeration which means that these people were enumerated in their usual residence. In this system a time reference of six months is used to determine whether a person who has been away would be enumerated at his usual residence or in the place where he is found during the census enumeration. If a person has been away from his place of usual residence for more than six months, he will be enumerated at the place where he is found by the enumerator even though that is not his place of usual residence. If he has been away from his usual place of residence less than six months he will be enumerated at his place of usual residence unless he has intentionally moved his place of usual residence in which case he would be enumerated at the new residence regardless of whether he had been there for six months or not.

For persons having no permanent residence such as homeless persons, crews of Indonesian ships sailing in Indonesian waters and mobile or floating houses, the 'de facto' method was used. the enumeration using this method was conducted simultaneously throughout the entire country.

The Census moment was fixed at midnight on September 24, 1971. The enumerators visited again all the households to see whether there were any changes due to birth or death occurring before midnight September 24, 1971, which was considered the Census day.

The control of the quality for the complete enumeration in the field was conducted by the supervisor. For each enumerator the supervisor selected one block which was not selected for the sample census and enumerated five households from that block listing only the names of the household members. The supervisor then compared his list with the enumeration conducted by the enumerator. If there were two or more differences, then the supervisor had to enumerate another five households and compare the results with those of the enumerator. If there were two, three, or four differences, the enumerator was considered acceptable and the supervisor would begin checking another enumerator, after advising the first enumerator of the existence of his errors. If there were five or more differences, the enumerator must re-enumerate the whole block. The differences were acceptable if they were caused by birth, death, or migration occurring during the period between the enumeration of the enumerator and the supervisor.

In order to complete their work the enumerators must submit summaries of each EB, which must be verified by the supervisor before submission to a higher level.

In West Irian since a population census had just been conducted in 1963, only the coastal regions were included in the Census and in the interior of West Irian the Census was conducted in a much simpler manner.

Sample Enumeration

The sample enumeration was conducted independently from the complete enumeration, so that the sample enumeration could be used as the post enumeration check of the complete census. The enumeration was begun on September 20 and ended October 4, 1971. Two kinds of forms were used, the household form and the individual form. The household form contained questions concerning the housing characteristics mentioned below:

  • number of occupants

  • type of ownership

  • occupancy of the building (wholly or partially)

  • number of rooms

  • construction material of outer walls

  • construction material of roof

  • construction material of floor

  • type of lighting

  • type of cooking fuel

  • source of drinking water

  • source of water for other usage

  • toilet facilities

  • garbage disposal facility

  • method disposing of garbage

This form was used more to collect the information concerning the household rather than to conduct a housing census.

The individual form contained questions asking detailed information from each member of the household on personal characteristics of:

  1. sex

  2. age

  3. relationship to head of household

  4. marital status

  5. citizenship

  6. religion

  7. language

  8. ability to read and write

  9. educational level completed

  10. field of education

  11. school attendance

  12. place of birth

  13. residence in other provinces

  14. last previous resident in another province

  15. total years lived in present province

For all married women and women who have ever been married, the individual form contained questions concerning:

  • number of children born alive

  • number of children still living

For those persons in the population 10 years old or over, the individual form also contained questions concerning economic characteristics:

  • type of economic activity for the last week

  • status (employer, employee, etc.)

  • occupation

  • industry

  • activity in agricultural sector in last season and his status

The sample enumerators were required to have higher qualifications and were trained more intensively as compared to the enumerators for the complete enumeration. There were 10,639 sample enumerators and 5,026 supervisors for a total of 16,551 persons. One enumerator covered one or two EB's and on the average one supervisor supervised two enumerators.

The training for these enumerators was given by census instructors who were trained at the central office in Jakarta. The training of the sample enumerators was planned to take place within the time span of three months (March through June 1971) and to finish before the July General Election. It was also planned that only a few instructors would be used to train all of the enumerators throughout the country by traveling from area to area to conduct the necessary training and in this way only two stages of training would be necessary. The fewer the training stages and instructors, the better in terms of ensuring the quality and consistency of the training. The more training stages used, the more likely it is that misunderstandings and misinterpretations will occur in the transmission of various concepts and definitions.

However, due to the Government regulation that no such meeting could be conducted within three months prior to the July1971 elections, the training schedule was compressed mainly within the month of August 1971: This meant that the number of first stage instructors would have to be increased far above the number originally planned and this also meant that the variation between the instruction would be greater which could also affect the quality of the enumeration.

There were 179 prospective instructors trained but only142 who passed the test and were actually sent to different parts of the country. Each enumerator received training for one week at the training centers which were set up in advance by the Provincial Census Directors and each instructor had to travel to two or three training centers in order that the training could be completed by the end of August 1971. The training included discussions of the concepts and definitions of all characteristics in the questionnaires, methods of interviewing, form handling, and a role playing was conducted in each class to assist the prospective enumerators in understanding the questions. A written test was given prior to going out to the field for a one-day field try out. Afterwards a discussion was conducted among the enumerators and instructors so that each enumerator could explain the difficulties he encountered during the field try out. The instructor could also point out any errors occurring in filling out the census forms.

The form handling training was one of the most important aspects of the training, since, as mentioned earlier, the optical readable forms were used for the sample enumeration. These forms would be used as a direct input to the computer and, therefore, certain requirements on the physical condition of the forms had to be met. In addition to form handling instructions, each enumerator was also supplied with a specially-designed plastic box in which to carry the forms during the enumeration in order to help guarantee that the forms, when returned to the central office in Jakarta, would be in an acceptable condition.

Publicity

Although the enumerators were intensively trained, the success of the Census could not be guaranteed unless the people being enumerated cooperated with the enumerators. To ensure this cooperation it was necessary to publicize the approaching Census; so this became a very important part of the Census preparation. Publicity programs were launched on the eve of the Census enumeration to educate the public on the need for the Census, the forms to be used, the kind of questions that would appear in the Census forms, and to assure the people that the information collected would be kept confidential and also to tell them of their legal obligation regarding the answering of the census questions.

All available media were used such as, radio, television, newspaper, magazines, film slides, pamphlets, etc. In addition, a Census song was made popular through radio and television and was taught at all schools in the country. School teachers, many of whom were also enumerators, were asked to help the Census campaign by telling their pupils about the importance of the Census so that the pupils could in turn tell their families. The village heads were also very cooperative in informing the public in their respective villages concerning the Census, so that at the time of the enumeration the people were well prepared to receive enumerators.

 Top of the Page

DATA PROCESSING AND TABULATION

The data processing activities started with the arrival of the Census documents from the various parts of the country at the central office and will end when all of the required tabulations have been printed.

Complete Enumeration Data

As was mentioned earlier, the processing of the result of the complete enumeration was done manually by using various summary forms that had been designed for this purpose. First, the enumerator made the EB's summary in duplicate. After being checked by the supervisor, one copy was sent to the central office and the other copy was kept at the village together with the original household forms.

 

The supervisor made the village summary to be submitted to the sub-district staff who in turn made the summary for the sub-district level. The regency or municipality staff made the summary for the regency or municipality level. Finally, the provincial staff made the summary for the entire province. These summaries included the tabulation of population by sex and age groups and also by citizenship.

The central office in Jakarta received copies of those summaries so that a checking for calculation errors could be conducted. The national total of population and households was obtained from this processing and was used as the universe for blowing up the population of household characteristics resulting from the sample enumeration.

Data Preparation

The processing of the sample census forms was done centrally at the central office. After finishing the enumeration and the necessary checking by the supervisor and the regional census staff, the sample census documents were sent directly to the central office.

Before the data could be processed by the computer, it was necessary to prepare the entire data manually at the 'Data Preparation Center' which was specially set up for this purpose.

The Data Preparation Center employed 200 persons who were trained in advance and consisted of six units: the receiving unit, the storage unit, the preliminary checking unit, the coding and marking unit, the batching unit, and the editing unit. A description of the duties of these units is as follows:

  1. The Receiving Unit - This unit was conveniently located for the incoming vans and trucks which carried the Census forms which were packed in specially-made heavy-duty cardboard boxes. This unit received the delivery slip, which contained a packing list, from the Provincial Census Directors. The unit then checked the contents of the boxes received for completeness. If there were any discrepancies, the unit chief must fill out a report form so that further clarification could be made by the Census administrator. These forms were then delivered to the Storage Unit.

  2. The Storage Unit - When the Census forms were received in this unit, they were transferred into pre-numbered plastic boxes according to the corresponding EB's code and then placed in a pre-determined and premarked slot in geographic sequence. The store keeper maintained a master register which showed the processing stage of each block in the store. The Storage Unit chief made a weekly report to the Steering committee concerning the progress of the data preparation which could be summarized from the master register.

  3. The Preliminary Checking Unit - Since the forms must be in an acceptable condition physically to be used as direct input to the computer, they had to be checked in advance. This checking included an inspection of the physical condition, checks for omissions, misplacement of the individual form, and the identification codes in addition a matching was made of the age of each individual with the date of birth which was written in the household form. The forms to be checked were ordered from the store and after the work for one day was finished, the forms were returned to the store. The movement of the forms was always controlled by means of a special ordering slip and was recorded on the master register kept by the Storage Unit.

  4. The Coding and Marking Unit - The occupation and industry of each person were completely written out in the questionnaire and were centrally coded by this unit. The unit leader ordered from the Storage Unit those EB's that had been checked and the coders would write in the codes for the occupation and industry according to the standard classification which had been prepared in advance based on ISCO and ISIC. The forms were then passed to a marker who marked the codes on the optical marking area of the questionnaire. A sample quality check was conducted to control the coding and marking errors. All forms were returned to the Storage Unit after the completion of work.

  5. The Batching Unit - This unit had the assignment of delivering the forms to the computer operational unit in batches. The unit had to first order the Storage Unit the forms which had been coded and marked and arrange them on special trolleys. One batch is more or less equal to the number of documents that could be read during one shift which is about 40 EB's. This unit was also responsible for the returning of the forms after being read by the computer.

  6. Editing Unit - While reading the forms the computer printed out error messages when it encountered the existence of non-permissible marks on the documents, missing documents (the number of individual forms must be equal to the number of individuals in a household as stated on the household form), unacceptable forms, etc. These error messages were sent to the Editing Unit where the corrections were made. The corrected forms were then reread.

Computer Processing

The computer system can be classified into three broad categories; these are

  • Transcription Phase

  • Edit and Imputation Phase

  • Tabulation Phase

  1. Transcription Phase - In this phase the documents were read by the Universal Document Transports (UDT) and the data on the forms transcribed onto magnetic tapes. Since so much effort had been put into keeping the forms in the best condition possible, the reading of the forms ran very smoothly and the number of rejected forms were negligible. The transcription of all the forms was completed in January 1973.

  2. Edit and Imputation Phase - The magnetic tapes containing data from the transcription phase were the input to this phase. Here the editing and imputation rules which had been designed during the preparatory stage were used to check the non response and the consistency of the characteristics in each household form as well as individual form. The editing and imputation rules were designed in such a way that the correction of one characteristic would only use the other characteristics from the same form and once a characteristic was corrected, then all other characteristics must be consistent with this characteristic. It was necessary, therefore, to establish priorities among all the characteristics on the form. As an example, the first priority on the individual form is the question on sex; the computer will check as follows:

Whether the question on sex is blank?

  • if no; accept as coded

  • if yes, ask:

Is the relationship to the head of household wife?

  • if yes, sex must be female

  • if no, ask:

Whether questions directed to ever married women are filled:

  • if yes, sex must be female

  • if no, ask:

Whether occupation is housewife?

  • if yes, sex must be female

  • if no, the sex would be determined by drawing a random number with certain probabilities assigned to male and female.

The next priority would be questions on children born alive and children still living which are only directed to the ever married women. In this case the sex would be considered as correct and would not be rechecked. This procedure continues until all of the characteristics have been checked. These rules were tested on a few provinces before being applied to all Census data.

The computer gives the imputation count for each characteristics within each regency or municipality. The level of imputation of each characteristic will be given in the publication so that the users will be aware of the quality of the data.

At present the editing of the data is still underway and is expected to be finished by May 1973.

3. Tabulation Phase - As soon as error-free data on magnetic tape was available the tabulation phase could be started. Computer programs, written and tested for this purpose, will tabulate the data according to the tabulation plans which were decided upon during the planning stage.

Due to the pressing need for information on some Census topics, a sub-sample of 10% from the total sample was drawn and processed in advance. This sub-sample processing was finished in June 1971 and was published in August 1971, as "Advance Tabulations". The tabulations of all data should be completed in August 1973.

Publications

The publications of the 1971 Population Census are made successively as soon as the processing of a particular tabulation is completed. The following are the publications which have been/will be produced from the 1971 Population Census:

I. SERIES A: Pre-Census publications. Tables resulting from household listing, classification codes, edits and imputations rules and the sample block codes.

  1. Series A no. 1: Number of blocks and households based on household listing. This publication is presented for each Kabupaten/Kotamadya (Regency/Municipality).
  2. Series A no. 2: Number of blocks and households in each Ketjamatan (Subdistrict) which are arranged in alphabetical order for each Province.
  3. Series A no. 3: Number of census buildings by type, based on household listing July - August 1970. This publication enables us to see the number of school buildings, mosques, churches, etc.
  4. Series A no. 4: Classification of Occupations and Industry for the purpose of the processing of the 1971 Population Census result. The classification presented in Series A no. 4 is constructed primarily for the processing of the census results and at the same time recognizing its international comparability. Therefore, this classification is arranged to conform with ISCO 1968 and ISIC 1968 and also in accordance with the conditions and requirements of the country.
  5. Series A no. 5: Edits and imputations for the the processing of 1971 Population Census results.
  6. Series A no. 6: List of Code Numbers and total selected census blocks for 1971 Population Census. The publication contains information on selected census blocks in each Regency/Municipality and the code number of each Regency/Municipality.

II. SERIES B: Preliminary tables containing tables on population by age, sex, and citizenship. This series is the result of complete enumeration.

1. Series B no. 1:

  1. Population by Province, sex and density.
  2. Population by Province, citizenship and sex.
  3. Population by Province, sex, and age (Urban, Rural and Urban + Rural).
  4. Population by Regency/Municipality, sex and age (Urban, rural and Urban + Rural).
  5. Percentage of population by Regency/Municipality, sex and age (Urban, Rural and Urban + Rural).

2. Series B no. 2: Contains data on population of Java-Madura by Sub-District, citizenship and sex.

3. Series B no. 3: Contains data on population of Outer Java-Madura by Sub-District, citizenship and sex.

III. SERIES C: Advance tables. This publication is the result of the sample census and contains tables based on some selected items from F2. This publication contains:

  1. Population by single years of age and sex.
  2. Population by age and sex.
  3. Population 5 years and over by literacy, region and sex.
  4. Population 10 years of age and over by age and over by educational attainment, region and sex..
  5. Population 5 years of age and over by school attendance, age, and sex.
  6. Population 10 years of age and over by age and type of activity.
  7. Economically active population by industry and sex.
  8. Economically active population by occupation and sex.
  9. Employed persons during the census week working in .......*) by status, region and sex.
  10. Unemployed persons during the census week over worked in .......*) by status, region and sex.
  11. Population working agriculture in the last season by status, region and sex.

*) each type of industry

IV. SERIES D: Population tables for all Indonesia. This series presents final tables on population showing the cross tabulations of the items in form F2 and contains:

  1. Population by single years of age and sex, Urban/Rural.
  2. Population by age and sex.
  3. Population by relation to head of household, Urban/Rural and sex.
  4. Male population by age and relation to head of household.
  5. Population by relation to head of household, sex and marital status.
  6. Head of household by marital status, age and sex.
  7. Male population by age and marital status.
  8. Male population by age and nationality.
  9. Population by religion, age and sex.
  10. Population by language, age and sex.
  11. Male population 5 years of age and over by literacy, region and sex.
  12. Population 5 years of age and over by literacy, region and sex.
  13. Male population by age and educational attainment.
  14. Male population by educational attainment and religion.
  15. Male population by educational attainment, age and field of education.
  16. Population 5 years of age and over by school attendance, age and sex.
  17. Female population by place of residence and place of birth. Population by place of residence and place of birth.
  18. Male population by place of residence and place of previous residence. Female population by place of residence and place of previous residence. Population by place of residence and place of previous residence.
  19. Ever married women by age and number of children born alive.
  20. Ever married women by age and number of children living.
  21. Ever married women by religion, age and average number of children born alive.
  22. Ever married women by religion, age and average number of children living.
  23. Ever married women by educational attainment, age and average number of children born alive.
  24. Ever married women by educational attainment, age and average number of children living.
  25. Ever married women by type of activity, age and average number of children born alive.
  26. Ever married women by type of activity, age and average number of children living.
  27. Female population by type of activity, marital status and age.
  28. Population by type of activity, Urban/Rural and sex.
  29. Male population by age and type of activity during the census week.
  30. Population by type of activity, citizenship and sex.
  31. Economically active population by status, Urban/Rural and sex. (31A). Unemployed persons by status, Urban/Rural and sex.
  32. Economically active population by status, age and sex. (32A). Unemployed persons by age, status and sex.
  33. Economically active population by occupation, age and sex. (33A). Unemployed persons by occupation, age and sex.
  34. Economically active population by occupation, educational attainment and sex. (34A). Unemployed persons by occupation, educational attainment and sex.
  35. Economically active population by occupation, status and sex. (35A). Unemployed persons by occupation, status, and sex.
  36. Economically active population by industry, Urban/Rural and sex. (36A). Unemployed persons by industry, Urban/Rural and sex.
  37. Economically active population by industry, age and sex. (37A). Unemployed persons by industry and sex.
  38. Economically active population by industry, educational attainment and sex. (38A). Unemployed persons by industry, educational attainment and sex.
  39. Economically active population by industry, status, and sex. (39A). Unemployed persons by industry, status and sex.
  40. Economically active population by occupation, industry and sex. (40A). Unemployed persons by occupation, industry and sex.

V. SERIES E: Population tables for each Province. This publication contains, in general, the same tables as for Indonesia given in Series D. These tables are by Regency/Municipality whereas the tables in Series D are given by Province.

VI. SERIES F: Housing tables for all Indonesia. This series shows the cross tabulation of the items which appear on form F1 and contains:

1A. Number of private households by Province and household size.

1B. Number of institutional households by Province and household size.

2A. Number of private households by number of rooms and household size.

2B. Number of institutional households by number of rooms and household size.

3A. Number of private households by Province and number of rooms.

3B. Number of institutional households by Province and number of rooms.

4A. Number of private households by Province, number of rooms and average number of persons per room.

5A. Number of private households by Province and tenure of living unit.

5B. Number of institutional households by Province and tenure of living unit.

6A. Number of private households by Province and type of living unit.

7A. Number of private households by Province and category of living unit.

7B. Number of institutional households by Province and category of living unit.

8A. Number of private households by number of rooms and category of living unit.

8B. Number of institutional households by number of rooms and category of living unit.

9A. Number of private households by category of facility and category of living unit.

10A. Number of private households having drinking water facility inside the yard by Province and type of water facility.

10B. Number of institutional households having drinking water facility inside the yard by Province and type of water facility.

11A. Number of private households having drinking water facility outside the yard by Province and type of water facility.

11B. Number of institutional households having drinking water facility outside the yard by Province and type of water facility.

12A. Number of private households by Province and type of water for other uses.

12B. Number of institutional households by Province and type of water for other use.

13A. Number of private households by type of drinking water facility and category of living unit.

13B. Number of institutional households by type of drinking water facility and category of living unit.

14A. Number of private households by Province and type of lighting.

14B. Number of institutional households by Province and type of lighting.

15A. Number of private households by type of lighting and category of living unit.

15B. Number of institutional households by type of lighting and category of living unit.

16A. Number of private households by type of fuel for cooking and category of living unit.

16B. Number of institutional households by Province and type of fuel for cooking.

17A. Number of private households by Province and type of fuel for cooking.

17B. Number of institutional households by Province and type of fuel for cooking.

18A. Number of private households by Province and type of toilet.

18B. Number of institutional households by Province and type of toilet.

19A. Number of private household by category of living unit and type of toilet.

19B. Number of institutional households by category of living unit and type of toilet.

20A. Number of private households by category of living unit and type of garbage disposal facility.

20B. Number of institutional households by category of living unit and type of garbage disposal facility.

21A. Number of private households by category of living unit and method of clearing out the garbage.

21B. Number of institutional households by category of living unit and method of clearing out the garbage.

VII. SERIES G: Housing tables for each Province. This publication contains, in general, the same tables as for Indonesia given in Series F, except that instead of classifying by Province, the tables are classified by Regency/Municipality.

VIII. SERIES H: Outline of 1971 Population Census results. This series contains analyses of results from the advance tabulation and comparisons with the results of the 1961 Population Census.

The above listing of publications is not exhaustive and more publications will be produced as more analyses are carried out.

 Top of the Page

EVALUATION AND ANALYSIS OF THE RESULTS

The total Indonesian population is the result of the complete enumeration in 1971 was 119,232,499. V.O. Kannisto(*) (* Report on mission to Indonesia, 14 December 1970 - 10 January 1971 United Nations, unpublished memograph.), before Census taking, had suggested that the correction of the complete count figures should be made using the result of the sample enumeration which served as the post-enumeration check. V.O. Kannisto realized that differences between the two counts may arise from the time lag between the operation; on the other hand, for differences not due to the time lag, he thought that it could not be considered that actual coverage errors occurred only or mainly in the complete count and that the sample enumeration is necessarily better. Therefore, he did not recommend to automatically adjust all data of the complete count to the findings of the sample enumeration, and suggested the following procedures:

  • (i) In each regency, compare the total of the sample count with the total of the complete count for the same blocks; if the difference is less than 2% on either side, accept the complete count.
  • (ii) If the sample count is more than 2% higher than the complete count, adjust the latter proportionately.
  • (iii) If the sample count is more than 2% lower than the complete count, examine whether the sample count in any census block is substantially (say 20%) lower than the complete count; eliminate such blocks from the comparison since they may indicate gross omissions, e.g. through boundary errors.
  • If after this scrutiny, in the remaining blocks as a whole, the sample count differs more than 2% from the complete count on either side, adjust the complete count proportionately either up or down as the case may now be; if the difference is less than 2%, accept the complete count.

Correction according to the above procedure has been conducted and the result was not different significantly.

Some demographers had suggested to always take the higher figure of the two enumerations with the assumption that in any population census, the probability of under enumeration is much higher than that of over enumeration.

It was decided however, that the population total resulting from the complete enumeration as stated above would be used as the total population for the base universe in blowing up the sample. The evaluation of this figure would be done separately by using the various methods available.

Comparison between those two enumerations are now being conducted on an individual basis by using the Chandra Sekar and Deming method.(*) (*) Chandra Sekar and W. Edward Deming, "On a Method of Estimating Birth and Death Rates and the Extent of Registration", Journal of the American Statistical Association, 44 (245) March 1949. This study was also meant to measure the level of under-enumeration of the census. Comparison will also be performed for certain characteristics.

Evaluation of the consistency of the data among characteristics was constantly conducted during the tabulation stage. This evaluation may result in the improvement of inadequate editing and imputation rules.

Although not all of the tabulations have been produced, analysis has been made of the available Census tabulations. A team of demographers within the Central Bureau of Statistics has been assigned to make the analysis and the following analysis has been given: age reporting in the Census, age pyramid, sex ratio, population growth, population projection, labor force, population densities, literacy, education, migration (internal), etc.

The Bureau has also invited Dr. N. Keyfitz from Harvard University and Dr. L.J. Cho and Dr. G. McNicoll from the East-West Population Institute to assist the Bureau in the evaluation and analysis of the 1971 Population Census data.

A collaborative project has been started in early 1972 between the Central Bureau of Statistics and the East-West Population Institute to study differential fertility of the population by using the "own children method". The technique required a special tabulation on the number of children living in the same household with their mother by the age of the mother. Another tabulation required is the number of children still living for the mortality adjustment of both children and women. The latter tabulations has been incorporated into the original tabulations plans, however, to produce the first tabulation a special effort had to be made as the question on own children was not specifically asked during the enumeration. It was very fortunate, however, that the interviewing procedure was done in such a way that it was still possible to relate the children (age 0-14) with their mothers. this was done as follows. For households which have only one mother, the relationship with the children can be directly made from the census tapes by a specially-written program. For households which have more than one mother, identification of the children of each mother in the household was more difficult. In this situation a manual processing was conducted from the census forms and then the result was combined with the first finding. This project is expected to be completed by the end of 1973.

Other analysis will be conducted as soon as the overall tabulations have been produced.

Top of the Page