developed by the United Nations (see UN [1956]) for projecting the growth of population. Projections using this technique usually rely on exogenous assumptions regarding changes in
fertility and mortality over some period of time. Such projections were first prepared for Papua New Guinea by Van de Kaa (1971) based on the assumptions set out in the
following table.
1 The discussion in this chapter is concerned only with the indigenous population of Papua New Guinea. The
model makes no attempt to project the age by sex composition of expatriate residents. The 50,000 expatriates enumerated in the 1971 Census of Population represented just over
2% of the total population. The number of expatriates has declined sharply since 1972 since a program for rapid indigenisation of the workforce was set in train, and this decline is expected to continue as fewer
Table 4.1; Assumptions of changes in fertility and mortality
Period
Life expectancy at birth in years
Changes in total fertility rate T.F.R. for 1961 is 6.225/woman used as index = 100
Males Females High Medium Low
1961-66 43.75 43.25 100.0 100.0 100.0 1966-71 48 .30 48.00 102.5 102.5 102.5 1971-76 52.35 52.75 105.0 105.0 100.0 1976-81 55.65 55.75 107.5 100.0 95.0 1981-86 58.20 58.50 105.0 95.0 90.0 1986-91 60.00 60.50 100.0 90.0 85.0
S o u r c e : Van de Kaa (1971), p.208 and p.213.
Van de Kaa's assumption of a rapid continuous
decline in mortality was based on the expected effectiveness of the increase in the quality and the coverage of public health services in Papua New Guinea. Van de Kaa also
predicted that after some time-lag, during which population would grow rapidly, the decline in mortality would be
followed by a fall in fertility rates.
Most projections of population growth for developing countries that have experienced a rapid fall in death rates in recent years have adopted similar assumptions for the time path of fertility. The process where an initial fall in death rates is followed after some delay by a corresponding
fall in birth rates has sometimes been termed the 'demographic transition'
Most of the writers on this subject agree that factors such as rising levels of education, rising incomes per head, increasing participation of women in the workforce,
a higher proportion of the population living in urban areas, or the promotion of birth control measures will be important
2
influences in reducing the birth rate. Van de Kaa suggested
the influence of urbanisation as the factor most likely to bring about an eventual fall in fertility in Papua New Guinea. However, no functional relationship between urbanisation and
fertility rates was proposed.
In the present model, as in the BACHUE series of
3
models, an attempt is made to link the projection of fertility rates to some of the factors cited above, taking advantage of the fact that changes in those factors are themselves
projected within the model. The relationships and parameters
used to project changes in fertility are set out in Section 4.3, and the basic projection of fertility rates and
population growth are compared to the projections derived by Van de Kaa.
1 See Coale (1973), 'The Demographic Transition
Reconsidered', International Union for the Scientific Study of Population, International Population Conference, Liege, 1973.
2 See Chapter 4, Volume I in U.N. (1973), The Determinants
and Consequences of Population Trends, Population Studies No.50, New York, 1973.
Section 4.2: The United Nations method for generating projections of population
This section describes the adapted form of the UN methodology used to project the growth of the indigenous
population of Papua New Guinea. Projections of population,
and its distribution by age and sex are based on assumptions about the time paths of fertility and mortality in rural
and urban areas. The derivation of these time paths in the
present model will be discussed in section 4.3.
Populations are projected over annual intervals
in two steps. The first is to estimate the survivors over
the interval, the second to estimate the number of births. West Model life tables'*' are used to estimate the ratio of
survivors from each age-sex group from either location. The model life tables give the ratio of survivors from one five year age group (e.g., males aged 15-19 years), to the next over a five-year interval, corresponding to a range of
values for the expectation of life at birth. The corresponding
survival ratios for any expectation of life at birth for either sex in either location are found by interpolation. These ratios are used to derive population projections over annual intervals as shown by the following example:
1 The selection of West Model life tables for projecting
the indigenous population of Papua New Guinea is
explained in Van de Kaa (1971), pp.117-21. The tables
Let r be the survival ratio, interpolated from the model life table for a given expectancy of life at birth such that, for example:
P (20-24, male, t + 5) = r.P(15-19, male, t) (4.1)
is the estimate for the 20-24 age group of males at t + 5. The estimates for the intervening years are set to be:
P (16-20, male, t + 1) = (r°-2).P (15-19, male, t)
P (17-21, male, t + 2) = (r°‘2) .P(16—20, male, t + 1)
etc.
In any year, the population in the conventionally used age groupings (15-19, 20-24 etc.), or in any single-year groups where required., can be calculated using Sprague multipliers.
The number of births over any annual interval is derived from estimates of total fertility rates and age- specific fertility rates in either location. These two measures are defined as follows:
The age-specific fertility rate (ASF (a^-a ) ) for any five-year age group from 15-19 to 45-49 is the number of live births per year to each woman in that age-group.
The total fertility rate (TFR) is the total number of live births that would occur to a woman who conformed exactly to the pattern of age-specific fertility rates throughout her child-bearing period
(15-49), so that:
TFR = 5 . [ASF(15-19) + ASF(20-24) + ...+ ASF(45-49)] (4.3) The model for Papua New Guinea derives changes in total fertility rates from the sub-model described in