Within the caveats in Section 3.1, a number of demographic studies show that population change does exert a strong first-order scaling effect on CO2 emissions models (O'Neill, 1996; Gaffin and O'Neill, 1997; O'Neill et al., 2000; Wexler, 1996). These studies support the notion that population growth and the policies that affect it are key factors for future emissions. Balanced against this, however, are other studies that take a more skeptical view (Kolsrud and Torrey, 1992; Birdsall, 1994; Preston, 1996). A full review of model results that address this question is given in O'Neill (1996) and Gaffin and O'Neill (1997) and is not be reproduced here because of space limitations (for a review see Gaffin, 1998). In essence, the controversy is one of the relationships between population growth and economic development, as well as other salient factors that influence emissions. These relationships were first discussed within the context of IPCC emissions scenarios by Alcamo et al. (1995) and are discussed in more detail in the following sections.
Figure 3-7 shows the long-established negative correlation between fertility rates and per capita income. Clearly, richer countries uniformly have a relatively low fertility rate. Poorer countries, on average, have a higher fertility rate. Lower fertility, however, does exist in some poor countries or regions, which illustrates the importance of social and institutional structures.
Barro (1997) reports a statistically significant correlation between per capita GDP growth and the variables life expectancy and fertility in his analysis of post-1960 growth performance of 100 countries. Other things being equal, growth rates correlate positively (higher) with increasing life expectancy and negatively (lower) with high fertility, which confirms the view that the affluent live longer and have fewer children.
Figure 3-7, a snapshot of many countries passing through the "demographic fertility transition" can be explained from both economic and socio-demographic points of view (Easterlin, 1978). Economically, Figure 3-7 can be interpreted as a reflection of the substitution that families make - away from having children and toward consuming more goods and services. With greater wealth, both goods and services become increasingly available as part of the families' "basket of choices" for consumption and, accordingly, they shift away from higher fertility to lower fertility. This move is further prompted by the rising relative costs of childcare, which include preferences that increase a child's quality of life, such as better schooling and extracurricular activities.
This income effect is primarily interpreted with respect to fertility changes in currently developing countries. In currently industrial countries, any change of fertility in response to increasing wealth is likely to be different from and probably even opposite to that of developing countries' fertility. Indeed, there is evidence of this in Eastern Europe, Sweden, Russia, and (recently) the United States (UN, 1997a), and it is linked to the question of long-term fertility rates in industrial countries (see Section 22.214.171.124.). However, it is evident that future world population levels will be dominated by growth in developing countries. Thus, if it is accepted that fertility is lower with greater affluence, then in emissions scenarios lower populations will still tend to correlate with higher per capita incomes.
Another countervailing factor is that mortality rates should also decline with wealth; as an isolated effect, this obviously results in higher population levels. However, the combined impact of both fertility and mortality reduction on population size is a net reduction in population levels (Lutz, 1996).
From a demographic point of view, the primary effect seen in Figure 3-7 is interpreted as infant and child mortality decline with increasing affluence. Families, in a sense, have to use birth control to achieve their desired number, which has always been lower than fertility rates. Increased affluence results in increased knowledge of, access to, and use of birth control, and accordingly families shift their reproductive behavior to lower fertility rates. Accompanying this basic premise is a host of complex social changes, including increased opportunities for education, employment, and non-maternal roles for women. Incorporation of the inverse relationship between economic and population growth in long-term emission scenarios is recent and mostly carried out in a qualitative way. Alcamo et al. (1995) reviews the literature available up to 1994. Nakicenovic et al. (1998a) report a long-term scenario study in which timing and extent of economic catch-up in developing countries were found to be tied to timing and pace of their demographic transition. Patterns of the range of per capita GDP growth rates for developing countries available in the literature (and given in Figure 3-10) also appear to reflect this relationship. Growth trends for the period 2020-2050 are generally higher than those for earlier or later periods; it is in this period that, according to demographic projections, the fastest change in demographic variables (especially fertility) will take place.
Other reports in this collection