M. Giovanna MerliCenter for Demography and EcologyUniversity of Wisconsin-Madison
Presented at the NICHD panel,"Visions of the Future: A Town Meeting on New Directions in Population Research"1Annual Meeting of the Population Association of AmericaMarch, 2000
What we are seeing today around the world are profound changes in the age structure of the population, and more rapid changes are foreseen in the decades to come. In developing regions, the largest-ever generation of young people is entering childbearing and working ages, the legacy of past high fertility and increased survival of infants and children. At the same time, in both developed and developing countries older populations are expanding. What is striking about the aging of the population of developing countries is its rapid growth and its large projected size. Age structure changes have broad implications for education and health of the young and for the social, medical and financial support for the elderly. The UNFPA State of the World Population Report for 1998 has given prominence to these "new generations," young adults and the elderly, focusing especially on the social and economic implications of these age structure changes in the developing world. During my year at the Rockefeller Foundation in 1998-99, the Population Sciences Division was evaluating the possibility of jumpstarting a new research initiative on "age structure changes and aging in developing countries." In fact, the participation in the initial exploration of these issues was part of my mandate there. Unfortunately, for reasons that had to do with a foundation-wide reorganization, these efforts waned.
The range of variation in age structure is obviously great and so is the nature of the demographic dynamics that produce these changes. Besides past and recent trends in mortality and fertility responsible for present and future population age distributions of developing countries, more sudden events such as high mortality associated with epidemics, wars and famines, regional displacements of individuals or rapidly occurring deficits in fertility are punching holes in the age structure of some populations leaving a long-lasting impact on communities and families. How communities and societies will be affected by these macro-changes depends on public policy as well as on individual responses. How do families cope with changes in age structure? How do they deal with many children, fewer children, with loss of adult workers, with aging? What are the short-term strategies adopted by families to accommodate to new conditions? What implications do these strategies have for the health and welfare of family members? Do these destabilizing events change the direction of intergenerational resource flows at the familial and societal levels? Does the adaptive behavior adopted by families feed back into the determinants of aggregate changes in the age structure?
These questions suggest a relationship between macro processes (the properties of the aggregate) and micro processes (those taking place among individuals), in particular how aggregate changes may affect individual decisions and how these in turn may feed back into systemic properties. The idea of examining the relationship between changes in a population age structure and individual decision-making, with feedbacks to the age structure is not new to the field of demography. In his 1978 PAA presidential address and in later writings, Easterlin (1978) examined the implications of changes in cohort size of the US working age population for labor force participation. He argued that increases in the size of young male cohorts relative to older cohorts weakened men’s labor market position, increased psychological stress and discouraged family building. Easterlin claimed that his analysis was consistent with the possibility of a self-generating mechanism producing repeated swings in the birth rate and age structure, and thus in socioeconomic conditions of a wide variety. The framework I propose here similarly incorporates individual responses to age structure changes and possible feedback mechanisms with respect to broad sociological themes such as the organization, transformation and inner relations of the family.
To shed light on these relationships, I first provide the example of one type of shock to the age structure with powerful implications for family relations and family reconfiguration. I am referring to the familial impact of HIV/AIDS morbidity and mortality in Sub-Saharan African countries most hard-hit by the epidemic. In particular, I focus on the indirect consequences of AIDS: the health status and the social and economic welfare of surviving family members, the short-term strategies adopted by families to accommodate the loss of their most productive members, and how these behaviors may trigger the emergence of a new stream of infections and new increases in prevalence. In the second part of this paper, I propose two additional examples of influences of characteristics of the aggregate system (age structure changes) on individual decision making with possible feedback mechanisms.
Most striking about the AIDS epidemic, especially if compared with other endemic diseases in the African and Asian regions is that it affects most severely the working age population and children (Quinn and others 1986). This age selectivity together with the disease’s long period of incubation and its low probability of survival has dramatic consequences for social organizations that are equal if not more powerful than those observed for other endemic diseases in Africa. The magnitude of age patterns of increases in AIDS mortality will have distinct consequences with implications for family relations, for the profile of well being of family members, while these in turn will provide mechanisms of feedback to the age structure.
The increased mortality of young parents will increase the incidence of orphanhood with consequences for children’s well-being. Orphans may be more likely to be malnourished and stunted than children who have parents to look after them. They may not receive the healthcare they need because it is believed they are infected with HIV and their illness untreatable. They may be the first to be denied education when extended families cannot afford to educate all the children. They may also find themselves with a huge burden of responsibility in their new acquired role of breadwinners for the family (UNAIDS 1999).
Although orphanhood should in theory affect male and female children alike, in practice this will depend on families’ coping strategies. Child fosterage, for example, could potentially absorb cost increases associated with the disruption caused by HIV/AIDS. But if fostered children experience higher mortality (Bledsoe and Brandon 1989) and there are strong female preference in fosterage arrangements (Page 1988), female children may end up experiencing the brunt of the crisis.However, this type of response may become increasingly untenable. With HIV/AIDS morbidity and mortality in the family, orphans become less attractive candidates for fosterage, their supply will exceed demand, and the disabling of the sending family will invalidate the principles of reciprocity involved in child fosterage, thus making this traditional arrangement less attractive and putting children at increased risk of child abandonment, school dropout and child-labor participation.
Such HIV/AIDS effects on orphanhood have clear implications for children’s well-being. But, perhaps more important, there are suggestions of possible feedback mechanisms into the age structure through the following mechanisms:
Because in much of Africa the aged are still intertwined in multigenerational living arrangements, most often with an adult child, increases in young adult mortality will decrease the availability of members of the intermediate generation, resulting in increasing familial relations between orphaned children and grandparents. Moreover, in the African context where multigenerational living arrangements facilitate the support of the elderly, the elderly suffer a double burden: They become caregivers of the younger generations, without the transfers from the middle generations, so that net resource flows may be from rather than to aging parents (Macwan'gi, Cliggett and Alter 1996; interviews in Kenya by Watkins, 1994), or, where resources are flowing from parents to children (Stecklov 1997), flows may intensify.
The effect of AIDS mortality and morbidity on the direction or level of resource flows may alter expectations regarding returns to investing in the schooling and other human resources of children, thereby leaving the young generation with alternative income earning options that may trigger the emergence of a new stream of infections and new increases in HIV prevalence.
The protracted nature of the disease, with its long health impairment, jeopardizes the family’s ability to generate resources for the care of children and of the elderly, thus increasing the social and psychic costs of those who are infected but also of those for whom the infected persons are in charge. In this case, damage to the family, in particular to orphans and the elderly, will start much before the time of death of one or both parents. This is what Palloni and Lee (1992:82) refer to as the "bottom of the iceberg, " in the sense that the effects of deterioration of the health status of adults on children is probably worse that the effect of their death.
Although these general mechanisms are clear, much remains to be done in the areas of both empirical estimation and forecasting. In recent work, Wachter, Knodel and Van Landingham (2000) (also see Van Landingham, Knodel, Im-em and Saengtienchai 1999) suggest ways in which the combination of demographic estimation techniques and micro simulation models can be used to assess the impact on kin availability for the elderly in countries such as Thailand that have been affected by HIV/AIDS.
To study the impact of HIV/AIDS on families, we need to specify choices for individuals depending on detailed circumstances. To this purpose, it may be easier to take individuals, rather than groups, as units of analysis thus making the use of microsimulations the preferred choice. Microsimulations are a tool to represent (a) complicated processes and the workings of changes in one factor through the systemic process, including feedback effects (Wachter 1987), and (b) a way of handling and incorporating the stochastic nature of individual behavior. They allow the representation of complex processes with many individuals and transitions from and to a limited number of states, thus making it possible to incorporate several dimensions in the model, i.e. kin availability, co-residence arrangements, health status of adults and duration in poor health, at the same time allowing one to separate the magnitude of the impact of changes in demographic conditions from the impact of HIV/AIDS mortality, or any other exogenous factor of interest, on the distribution of population by kin types.
An additional promising line of inquiry into the familial impact of HIV/AIDS mortality and morbidity would be the application of an accounting framework developed by Lee (1994, 1995), which allows one to measure the aggregate allocation of wealth between age groups and the direction of intergenerational resource flows, to available data from Sub-Saharan African countries with high HIV prevalence. Stecklov (1997) applied this framework to 1986 Côte d’Ivoire data, collected for the World Bank’s Living Standards Measurement Survey, to test Caldwell’s theory of intergenerational wealth flows and fertility. His findings counter Caldwell’s predictions, and suggest that wealth flows in Cote d’Ivoire, a high fertility country, run downward from parents to children. His conclusion is that, despite their high costs, children still remain parents’ best source of old age support. Thus, an interesting question would be the extent to which the levels and direction of intergenerational transfers are "aggravated" by the familial effects of HIV/AIDS. 2
Although the major emphasis of this brief presentation is on the familial impact of HIV/AIDS mortality and morbidity, the proposed lines of inquiry can be similarly applied to examine the familial impact of other diseases of epidemic proportions ( e.g. malaria), of wars, and of other contemporary processes that drill holes in the age structure and significantly alter traditional family relations such as urbanization, rural-urban migration (China, Africa), and international migration (Mexico, Africa).
To these more sudden perturbations in the age structure, I would add the very rapid fertility decline over the last three decades in China, its long term implications for the welfare of the elderly, and the possibility of a self-generating mechanism which would boost fertility up again. The Chinese TFR has declined from 6 in the early 1970s to the current 1.8, the result of socioeconomic change but especially of a succession of draconian population policies. Together with a dramatic mortality decline and the large baby-boom cohorts born in the 1950s, 1960s, and 1970s, this rapid fertility decline forms the demographic base for the extremely rapid aging of the Chinese population. The proportion of the population 60+ is now 10%, but is project to increase to 30% in 2050 according to the medium fertility variant of the UN projections as revised in1998.
What would it take for fertility to increase to balance off the aging of the huge Chinese population? To most, this question may sound naive, as there seems to be only one possible answer. Although the Chinese population policy has been successful in changing reproductive behavior to levels below what people really want, the persistence of preferences that cluster around two children, one of each sex, means that the current family planning policy and its implementation have not been sufficiently successful in squelching a latent demand for children, especially in rural areas. Thus, relaxation of the policy would soon result in fertility increase.
Yet, there are signs that current levels of fertility may continue even in the absence of the policy that limits urban couples to 1 child and rural couples to 1.5 children (couples with one daughter are allowed a second child). The birth planning policy has been in place at a time when even rural China has been strongly affected by massive liberalization in the economic realm, leading to a greater commodification and monetization of everyday life—all factors implicated worldwide in the switchover from high to low fertility (Caldwell 1976; Freedman 1979; Mason 1997). Moreover, with a rapidly changing opportunity structure introduced by sweeping economic changes in urban areas, the calculus of material advantage may occupy a more prominent place in decisions about family formation in the near future. To these structural changes and conditions should be added the very persistence of the One-Child Policy and its rural adaptations. It is now in its third decade. A generation of young people has grown up in its thrall. Memory of alternative choices and policies must be fading and the possibility that compliance with the policy may be in the process of shading into acceptance should not be dismissed out of hand (Merli and Smith 2000). Moreover, there is enough evidence to suggest that in urban areas, and to a certain extent in rural areas as well, there is a growing sense that a daughter could help support her parents in old age better than a son (Greenhalgh, Zhu and Li 1994; Zeng and George 2000). If, in the long run, this family arrangement were to become a common preference for old age support, the absolute importance of a bearing a son to guarantee support in old age, now the main factor keeping rural fertility above government’s goals, would loose its basis of existence. Fertility responses needed to balance the aging of the population, equivalent to an estimated TFR of 2.5 (Zeng and George 2000), may be more difficult to obtain than previously thought, even in the absence of the policy.
This would contribute to an inexorable rapidity of population aging. But an increasingly realistic awareness of the aging of the population, augmented by concerns about ways to ensure old age support, may influence fertility motivations and trigger a surge in fertility. An important point here is that individual behavior does not only respond to financial considerations, but also to demographic realities and to the perceptions and ideas these generate.
Perhaps a more telling example is provided by the Italian case. Italy’s TFR is currently 1.2; 24% of its population is over age 60; and this figure is projected to rise to 41% by 2050. 3 According to UN estimates, if Italy wishes to maintain the current ratio of four persons of working age for every retired person (65+), it will need an average number of migrants over the period 1995-2050 of 2.2 million per year (UN 2000). The Italian press is already portraying the Italian population in the middle of this century as an old population with a labor force mostly made up of recent immigrants or children of immigrants from the Balkans and North African countries. This awareness is forming at a time when replacement migration is seen in international policy circles as a possible remedy for the decline and aging of European populations (UN 2000). How long will it take for the prospects of an aging Italian-born population supported mainly by a young population of immigrants to sink into the collective consciousness of Italians, perhaps assisted in this formulation by some rambunctious political figure of dubious ideological stature, a domestic version of Austria’s Joerg Haider or an international version of the leader of Italy’s Northern League, Umberto Bossi, and to revert to fertility increases as a way of preventing this from happening?
1 Thanks are due to Alberto Palloni and Elizabeth Thomson for their suggestions and comments.
2 In Côte d'Ivoire, the HIV adult rate of infection has increased from almost negligible levels in the mid-1980s to the current rate of 10%. Since 1986, three additional rounds of LSMS have been conducted. LSMS data are also available for Tanzania, a high HIV prevalence country.
3 Projections are according to the UN medium fertility variance with some allowance for in-migration.
Bledsoe, C. and A. Brandon. 1989. "Le placement des enfants et son influence sur la mortalité," in G. Pison, E. Van de Walle and M. Sala-Diakanda eds. Mortalité et Société en Afrique, Paris: Presses Universitaires de France, for INED.
Caldwell, John C. 1976. "Toward a restatement of demographic transition theory." Population and Development Review 2(3/4):321-366.
Easterlin, R. A. 1978. "What will 1984 be like? Socioeconomic implications of recent twists in age structure" Demography 15(4):397-432.
Freedman, Ronald. 1979. "Theories of fertility decline: A reappraisal." Social Forces 58:1-17.
Greenhalgh, Susan, Zhu Chuzhu, and Li Nan. 1994. "Restraining population growth in three Chinese villages, 1988-93." Population and Development Review 20(2): 365-395.
Lee, R.D. 1994. "Fertility, mortality, and intergenerational transfers: Comparisons across steady states," in J. Ermish and N. Ogawa, eds., The Family, the Market, and the State in Ageing Societies. New York: Oxford University Press.
------. 1995. "Population age structure, intergenerational transfers, and wealth: A new approach, with applications to the Unites States," Journal of Human Resources 29(4):1027-1063.
Mason, Karen Oppenheim. 1997. "Explaining fertility transitions." Demography 34(4):443-454.
Macwan'gi, M. L. Cliggett, and G. Alter. 1996. "Consequences of rural-urban migration on support for the elderly in Zambia," Presented at the Annual Meeting of the Population Association of America, New Orleans, Louisiana, May 9-11, 1996.
Merli, M. G. and H.L. Smith. 2000. "Has the Chinese birth planning program been successful in changing fertility preferences? Evidence from linked records in four counties in rural northern China," Paper prepared for presentation at the session "Fertility in Asia," PAA Annual Meetings, Los Angeles, March 23-25, 2000.
Page, H. 1988. "Childrearing versus childbearing: Coresidence of mother and child in Sub-Saharan Africa," in R. Lesthaeghe, ed. Reproduction and Social Organization in Sub-Saharan Africa Berkeley: University of California Press.
Palloni, A. and Y.J. Lee. 1992. "Some aspects of the social context of HIV and its effects on women, children and families" Population Bulletin of the United Nations 33:64-87.
Quinn T.C; J.M. Mann, J.W. Curran, P. Piot. 1986. "AIDS in Africa: An epidemiologic paradigm" Science 234(4779):955-63.
Stecklov, G. 1997. "Intergenerational resource flows in Côte d’Ivoire" Population and Development Review 23(3):525-553.
UNAIDS. 1999. Children Orphaned by AIDS. Geneva: UNAIDS.
United Nations. Population Division. 2000. "Replacement migration: Is it a solution to declining and ageing populations?" Press document. January 6, 2000.
VanLandingham, M., J. Knodel, W. Im-em, and C. Saengtienchai. 1999. "The impact of HIV/AIDS on older populations in developing countries: Some observations based upon the Thai case. Research Report No. 99-441. Population Studies Center, University of Michigan.
Wachter, K., J. Knodel and M. VanLandingham. 2000. "AIDS and the elderly in Thailand: Projecting familial impacts" Unpublished manuscript. University of California, Berkeley.
Wachter, K. 1987. "Microsimulation of Household Cycles," in J. Bongaarts, T. Burch and K.
Wachter, eds. Family Demography. Oxford: Clarendon Press.
Zeng, Y. and L. George. 2000. "Extremely rapid ageing and the living arrangements of older persons : the case of China." Paper presented at the United Nations Technical Meeting on Population Ageing and Living Arrangements of Older Persons: Critical Issues and Policy Responses, New York, February 8-10 2000.