From this constant percentage of forty for the non-adults of the population Wertheim deduced the correlation with the birthrate of such a population, and the correlation with the age-gradation within the non-adult group (0-14-years-old). It is this last Kind of correlation that we shall now go into. The age-gradation of the non-adult group will be closely connected with the death rate of infants, toddlers and children at the school-going age. This death rate is unknown in countries in rapid social change because of a lack of civil registration. A high death rate among infants and toddlers will require a higher birthrate to maintain the non-adult group at 40% of the total population than when this death rate is low. A population-pyramid will have a wider base and rise more gradually when the death rate is high.
Starting out from three propositions, which are very probably also consistent with the real conditions of the population in the D.I. Jogjakarta, Wertheim drew up three death-rate-tables based on data obtained from the investigations made by Haas and Brand in Indonesia before the last world-war and from Smith’s data about Malaya of 1947. The three propositions are the following:
a) The birth-rate is almost constant.
b) The annual natural increase of the population is 2%.
c) A rural population is considered, where the immigration factor does not materially influence age-gradation.
The first table started out from a death-rate among infants of 6%, which was looked upon as a minimum value for a population with 40% non-adults, This figure was chosen in view of de Haas’s researches into the death-rate within the European population group in Djakarta in 1935-1936 where a death-rate of 5.6% was found among infants. Data about death among the European group in Indonesia in the years 1929 and 1931 point to a death-rate among male infants of 7.43% and among female infants of 5.8% But then this group lived under very favourable circumstances at the time. The third table started out from a death-rate of 25% among infants, which was looked upon as the upper limit. This figure was derived from de Haas’s researches with this restriction, however, that in de Haas’s computation of the death-rate among infants of the Indonesian population of Djakarta in 1935-1936, he started out from a birth-rate of 35%, and a death-rate of 30%. A conversion of his data, when starting out from a birth-rate of 40% , results in a death-rate among infants of 25.7% The second table starts out from a death-rate of infants of 15% as an intermediate value. In view of the data about the death-rate of infants in Jogjakarta (1952-1956), we estimate this at about 15%. So this deduced value was chosen for our guidance in computing the size of the 0-15 years-old group.
The three age-tables computed by Wertheim are shown in Table III-6 together with the data of the age-pyramids of the groups of Manahassers, Sangiresians, Talauds, and Ambonesians as they appeared from the 1930 census. This table III-6 also gives the results of the surveys made in Java in the period 1955-1958. These surveys are of Senen (Djakarta), of Tanah Tinggi (Djakarta), of the village of Djabres (Central-Java), of the village of Andir (West-Java), of the two desas near Tjikarang (West-Java), and of the district Kraton in the city of Jogjakarta. When comparing the series of figures of the 1930 census with those of Senensurvey of 1956, we see a dip in all the 5 series in the groups of the two-years-old. Irregularities in age-gradation, in the sense that one year-group is larger than a younger on, are found in all these series in three of four age-groups. In all the series a group of ten-years-old is to be noticed, resulting in a second dip for the eleven-years-old, so that the group of twelve-years-old is larger than that of the eleven-years-old in all the 5 series. This is caused by the tendency to round off ages to fives and tens noticed in several cases, when a census took place in Asia and South-America. The dip in the two-years-old group was also found in Malaya in the 1947 census in the case of Malayans. Malayans together with their children that had immigrated, mostly of Chinese descent were called “Malaysians” and left out of consideration. Age-groups of 0-1-2-3-4 and 5-9-years-old were distinguished. In the Malayan states, too, this irregularity was found in the group of the on-and tow-years-old notwithstanding a proper civil registration and high percentages of literates.
Neither does a high percentage of literates appear to be a full guarantee for obtaining a reliable age-pyramid of the group of young children. The registration of births being quite good, no deficiency of this could account for the phenomenon. The birth-rates of Malay-states ranged from 38.1 to 47.1 per 1000 inhabitants, and can therefore be called high. Smith presumed in his “Population growth in Malaya” that incorrect statements of ages might explain this irregularity, Comparison of the figures obtained in the count with the figures of the birth and death-rates, made it seem probable that the number of children counted should be correct, whereas the statement of ages could not be. A systematic “overstatement” of ages in the case of infants and toddlers in such a way that parents think they are one year older after a period shorter than one calender-year seems likely e.g.:
Those whose correct age is 0 -10 months are reported 0 years
” ” ” ” ” 11-17 ” ” ” 1 years
” ” ” ” ” 18-29 ” ” ” 2 years
” ” ” ” ” 30-41 ” ” ” 3 years
The total number of children up to five years of age being correct, this results in a wrong distribution of the age-groups. Smith gave the following explanation: “The phenomenon of a dip at age 1 in the graph of the single year age of children is a common one in the Muslim world and is to be attributed to the fact that the birthday is not a remembered date in the life of a Muslim. After a child has been weaned and its age is no longer counted in months, age will be determined in relation to an important event in the year, possibly the rice-harvest in rice-producing areas, or an important date of dates in the Muslim year in other areas”. As we found this dip in the on-year group together with the other irregularities in the four Indonesian population-groups, which largely profess Christianity and use the Christian era, any connection with the Islam can be doubted. To find out if gradation in the pre-school-age period of Children in the D.I. Jogjakarta could be approximately established, we adapted the data of the Central Health office for combating endemic diseases, (Chief Dr. R. Kodyat). This service organises the cure of yaws in all the ketjamatans of Java. To this end it sets about as follows: The preparatory work comprises a request to the lurah desa for a copy of the civil-registration of the desa, as full and recent as possible. This list of all the residents is used in the first clinical survey to find out whether every one was examined for symptoms of yaws On receipt of the list the survey is made a few weeks later. Thus it appeared that in two unselected ketjamatans Godean, (Sleman) and Panggang, (Gunung Kidul) 61% and 92.5% respectively of all the residents were examined for signs of yaws in the first survey. Patients discovered are treated, and in the course of years, new surveys are made, until a satisfying situation is reached. List of residents of all the dukuhs of the ketjamatans Godean, and Panggang, containing 77, and 74 dukuhs respectively, comprise data about names, sex, and age, data from some weeks before the survey. From these lists we arranged the residents in the following 10 separate age-groups: viz.0-, 1-, 2-, 3-, 4-, 5- and 6 years of age, 7-9 years, 10-14 years,15 years and over. The result of this arrangement is shown as far as the group 0-14 years is concerned in table III-6.
Table III – 6.
Division of the non-adults over the different age-groups concerning several population-groups in Indonesia, compared with Wertheim’s tables according to the 40% test,expressed in percentage of the total population
Of these data some irregularities are very striking. Viz. The large group of one-years-old and the small group of 0-years-old. In the ketjamatan Panggang the group of three-yeas-old appears to be larger than that of the two-years-old, and the group of five-years-old larger than that of four-years-old. In Godean the group of six-years-old is larger than that of five-years-old. In both ketjamatans the groups of ten-to fourteen-years-old are larger than those of the five-to nine-years-old, which may be explained form the fact that the five-to nine-years-old were born in 1941-1946 during the Japanese occupation, and may have been subject to a greater mortality among infants and toddlers than the ten-to fourteen-years-old. In surveys made in West-Java in 1956-1957 the group of ten-to fourteen-years-old appeared also to be relatively small. This concerned also the group of children born in 1941-1946. Because of the many irregularities these data did not seem to use to be very suitable for the purposes we wanted them for, but we did find some striking features:
This being a house-to-house registration, and the houses having numbers, it seems improbable that it should have been desultory. In the first survey, where the persons examined were controlled with the registration-list, it appeared to be necessary to make additions all the time. The number of people who were not registered, but who were examined by the official of the service for the prevention of yaws, and who were also added to the list was considerably larger than that of the people on the list who had died or moved to another dukuh and were consequently crossed out in the list. It appeared that 4 additions were made against one removal. From these corrections it seems likely that the number of actual residents is higher than the number of residents registered. In the ketjamatans Godean, and Panggang in 38%, and 13% of the dukuhs respectively only two, one or no babies were found in the civil registration, which is improbably low for an average of 400-500 residents per dukuh. This must be explained from an imperfect registration rather than from anything else. The number of new residents added to the list and examined during the survey, but who did not occur in the registration list received from the lurah consisted largely of children. Of 100 residents added to the list during the survey 9 were infants, 35 one-to four-years-old toddlers, and 27 five-to fourteen-years-old children; in 29 cases children of 15 and older people. So this affects the children’s group most. All these corrections made during the survey lead us to suspect that under-reporting in the desa-registration particularly affected the younger year-groups, and naturally all the new-born children.
Corrections because of deaths concerned in 45% of the cases children under five, in 52% children under fifteen that had to be accounted for in the registration. So in the younger age-groups the number of mutations is largest, and involves most work in keeping the administration up to date. That under-reporting should be greatest in these age-groups is therefore not incredible. The degree of incompleteness is difficult to assess. The total number of non-adults in Godean and Panggang amounted to 39%, and 42% of the total population respectively. These registration data are too irregular to serve for our guidance as to the size of the various age-groups of the population of the D.I. Jogjakarta. As even accurate census-and survey-data give an irregular gradation of ages among the under-fifteens, the real gradation can only be found when there is a complete civil registration accurately posted up.
As this does not yet exits in the D.I. Jogjakarta, the only data usable for us are Wertheim’s computations.
The second life table, starting out from a death-rate among infants of 15% supplies us for the various kabupatens and for the city of Jogjakarta with the flowing figures for our guidance in the estimate of child mortality in the different year-groups (Cf.Chapter VI). Which we hope will not be very far wrong.
Table III – 7
Age-composition of the non-adults in the D.I. Jogjakarta in 1955 (Dec.)
calculated according to Wertheim’s 40% test, and supposing
an infant mortality of 15% and rapid growing population.
