The Network of Intermediate Model Stable Populations

A solution to the problem posed by the very title of this work was given when, in chapter V, we established that a population having an age distribl;ltion which remains invariable with the passage of time can be assimilated at all times to a Malthusian population (we gave the name of "semi-Malthusian" to these populations with constant age distribution). Once we verify that an actual population is semiMalthusian, as most populations of developing count~ies are, all the methods studied in connexion with Malthusian populations, with the aim of determining the chara;ct~r­ istics of a population when some of those characteristics are known become applicable. As has already been mentioned 'more than once, however, the invariability of the age distribution of many actual P?pulations could be interpreted in a different manner; instead of assimilating them to semi-Malthusian populations, one could assimilate them to quasi-stable populations. The concept of quasi-stable populations was introduced into demography for the first time in the United Nations study on world population prospects,' where models of demographic evolution were constructed for the purpose of estimating the future population of the various countries of the world. These numerical calculations showed, among other things, that a population with invariable fertility but with mortality varying within the universe of the intermediate series of model life tables also had an almost invariable age distribution. That result was given more general application; a quasi-stable population was defined as a population in which fertility hadfor long been constant and the age distribution also remained more or less constant. The concept of quasi-stable populations is therefore based on experience, and it was defined as a result .of numerical calculations carried out on a necessarily limited scale. It is therefore not impossible that other calculations made with other universes of model life tables would result in a modification of the definition of this concept. For the time being, however, we shall take this concept exactly as it has been defined. It is easy to see how it can be used to solve the problem pose~ ?y the title of the present work. It is known that the fertihty of most of the developingcountries has remained constant, or at least has varied only slightly, for a long time past. Their populations are therefore quasi-st~bl~ populati~~s, and it can be verified that their age distribution vanes little with the passage of time. Let us now imagine that we have available a network of quasi-stable populations, computed once for all by