Population Dynamics of Senegal (1995)

Over the last two decades, child mortality has declined rapidly in Senegal, while changes in fertility have been small. According to the 1992-1993 DHS-II, the total fertility rate in the 4 years before the survey was 6.1, while the probability of dying by age 5 in the 5 years before the survey was 131 per 1,000 live births. These two figures imply a rate of natural increase of close to 3 percent per year, and since we find no evidence of major net migration, the actual rate of population growth is also likely to be close to 3 percent. This rapid growth can be attributed to both sustained high fertility—on average each woman bears approximately six children—and declining mortality. What are the prospects for future change?

Two recent reviews of fertility transition in Asia and Latin America underscore the point that no unique characteristic or combination of characteristics represents necessary and sufficient conditions for substantial fertility decline in a population over the course of economic development (Casterline, 1994; Guzman, 1994). However, several elements appear particularly important: low child mortality, high female education, well-organized national family planning programs, and the existence of a desire to have fewer children.

In Senegal, child mortality is falling rapidly, but it is not yet low;

female education remains low; family-size desires are falling quite rapidly, but they are still close to actual fertility; and the family planning program is quite weak. In 1982, family planning efforts in Botswana, Kenya, Senegal, and Zimbabwe were all quite weak, and the Ross-Mauldin Index¹ for all countries was very similar: between 23 and 28 percent of the maximum score. By 1989, however, the family planning programs in Kenya and Zimbabwe had increased in strength to between 56 and 58 percent of the maximum score, and the program in Botswana had increased in strength to 75 percent of the maximum score, making it among the strongest programs in the developing world. In Senegal, the score had increased some what—to 44 percent of the maximum—but not enough to avoid classification as a weak program.

Not surprisingly, the decline in fertility in Senegal has been quite modest: national demographic surveys indicate that fertility declined about one child per woman between 1975-1978 and 1989-1992, and most of that decline took place among urban women. The driving force behind these changes appears to have been a trend towards later marriage and later first birth, and the decline in fertility occurred almost entirely among women under age 30. Little of the fertility decline in Senegal appears to be attributable to either a decrease in ideal family size or an increase in the use of modern contraception, although the proportion of women using modern contraception has increased over the very recent past. Current preferences for number of children still lie very close to the physiological maximum level, given a continued regime of delayed marriage and long birth intervals.

Assuming Senegal achieves further increases in primary and secondary school enrollment for women, as well as greater urbanization, further fertility declines can be expected to occur in the near future. In rural areas, further declines in actual fertility can still be achieved through the mechanism of later marriage. However, in urban areas, particularly Dakar, most of the decline in actual fertility that is achievable solely by an increase in age at marriage has already occurred, so that future fertility reductions must await greater coverage of modern contraception.

The World Development Report 1993: Investing in Health (World Bank, 1993b) identifies three key factors in the dramatic mortality declines of the twentieth century: income growth; improvements in medical technology; and the implementation of public health measures, including both infrastructure and the spread of knowledge. Income growth allows the purchase of more food, better housing, and more health care. Advances in medical technology have increased the amount of health improvement achievable through a given expenditure on health, although they can have a beneficial

effect only among a population with access to services. Public health infrastructure improvements, including primary health care services, as well as water supply and sanitation, have reduced exposure to pathogens and contributed to disease prevention. The spread of knowledge about disease prevention and treatment has also contributed greatly to mortality decline worldwide, and has been accelerated by rising levels of education, particularly among mothers.

Senegal has experienced particularly rapid declines in child mortality over the last two decades. The probability of dying by age 5 (per 1,000 live births) declined from about 265 in 1975 to about 135 in 1990. Child mortality decline started in urban areas in the 1950s, but did not spread to all rural areas until the late 1970s. Income gains in Senegal over the last two or three decades have been negligible, as documented in Chapter 2, and can have contributed little to the mortality decline. Nor can improvements in medical technology explain the sudden decline in child mortality in the mid-1970s. There are large differentials in child mortality by education of mother in Senegal, though overall educational levels remain low, and improvement has been quite slow: 73 percent of women of reproductive age had no education in the 1992-1993 DHS-II, as compared with 77 percent in the 1986 DHS-I. The decline in child mortality in Senegal, at least in rural areas, appears to be most closely linked to the improved health infrastructure, particularly the primary health care policy introduced in 1978 and the implementation of the World Health Organization Expanded Programme on Immunization starting in 1981. The experience of one rural area, Mlomp in southern Senegal, provides support for this conclusion. Mortality decline in Mlomp, well documented by a population observatory, was both early and very rapid, and appears to have resulted from the establishment of a private dispensary and maternity clinic in the area. Thus of the three key factors in mortality decline identified by the World Development Report, improvement in the health infrastructure, particularly the provision of primary health care, appears to have been the most important in Senegal's mortality decline.

The above picture of demographic change suggests that Senegal's small fertility decline is unlike the recent declines in other sub-Saharan African countries. In Botswana, Kenya, and Zimbabwe—usually considered the three countries in the vanguard of African fertility transition—fertility declines have been associated with increased use of modern contraception. Botswana, Kenya, and, to a lesser extent, Zimbabwe have also experienced a decline in teenage marriages, but fertility rates under age 20 have not

changed much. Consequently, these countries have experienced an increase in nonmarital fertility. This separation of marriage and fertility has had the effect of reducing the influence of marriage on fertility rates (Working Group on the Social Dynamics of Adolescent Fertility, 1993). In Senegal, fertility has fallen because marriage has been delayed, and marriage and fertility have remained linked.

The Senegalese pattern matches more closely, though not exactly, the pattern found in certain northern African countries during the first phase of their fertility declines. For example, most of the initial decline in fertility observed in countries such as Algeria, Egypt, and Tunisia can be attributed to later age at first marriage (Fargues, 1989; National Research Council, 1982). In these countries, the initial phase of fertility decline was immediately followed by a second phase linked to a substantial decline in the demand for children and a corresponding increase in use of modern contraception among married women.

Though we cannot know for certain what the demographic future of Senegal will be, it may be revealing to compare Senegal with another country that is further along in its demographic transition to see what may lie ahead. Because the emerging pattern of fertility change in Senegal appears to be more similar to the changes that have occurred in the countries of northern Africa than to those that have occurred in sub-Saharan Africa, it may be useful to compare Senegal with a northern African country. For the sake of discussion, we have selected Tunisia, the country of northern Africa that has progressed furthest towards reduced fertility and child mortality.

A comparison of the decline in child mortality in the two countries indicates that the rate of decline in Senegal after 1975 is very similar to that in Tunisia from 1960 on. In Tunisia, mortality before age 5 declined steadily from about 250 per 1,000 live births in 1960 to around 100 in 1980, to around 45 in 1990 (Hill and Yazbeck, 1994). Using the same trend-fitting methodology, mortality before age 5 in Senegal declined from about 300 in 1960 to about 265 in 1975, to about 135 in 1990. Thus, Senegal is currently following a child mortality pattern similar to that followed about 15 years earlier by Tunisia.

The parallels between the subsequent fertility declines in the two countries are also clear, though the time lag is greater—close to 20 years.² In Tunisia, total fertility (between the ages of 15 and 35) was about 5.3 in the early 1960s. It fell to 4.2 by the mid-1970s and to about 3.4 by the mid-1980s. In Senegal, total fertility (ages 15-35), which had not changed appreciably since the 1960s, was 5.3 in the mid-1980s, a figure closely resembling that in Tunisia 20 years earlier. In 1991, total fertility in Senegal had fallen to 4.5³ —a level comparable to that in Tunisia 16 years earlier.

The initial decline in fertility in northern Africa, as in Senegal, can be attributed to a later age at first marriage and first birth: in northern Africa,

this initial decline was immediately followed by a second stage, characterized by falling ideal family size and rising contraceptive use. An analysis, using Senegalese data, of the variables associated with Tunisia's continued fertility decline may allow speculation about fertility change in Senegal in the future.

Fertility differentials may be harbingers of fertility trends, and differentials in median age at first birth in Senegal suggest the potential for rapid change. For women aged 25-29 in 1992-1993, the median age at first birth for urban women was 20.8, compared with 18.5 in rural areas. The contrast was even more pronounced by education: 24.5 years for women with secondary or higher education versus 18.7 years for women with no education. In Tunisia in 1988, the urban-rural medians were 25.1 and 23.5, respectively, and the secondary+ education versus no education medians were 25.7 and 23.4, respectively. Thus for well-educated women, patterns of entry into childbearing are not very different in Senegal than in Tunisia, though the differences among less-educated women are very large.

Differences between actual fertility and fertility preferences should also provide insights into future change. At the time of the Tunisia WFS in 1978, total fertility (ages 15-49) was 5.9, but desired family size was only 4.2, and rate of use of modern contraceptives among married women was 25 percent. By 1988, total fertility had dropped to 4.4, desired family size had fallen to 3.5, and rate of use of modern contraception had increased to 40 percent. In Senegal, total fertility in 1978 was 7.1, while desired family size was 8.3, and rate of use of modern contraception was less than 1 percent. By 1986, total fertility reportedly had fallen to 6.6, desired family size had fallen to 6.8, and rate of use of modern contraception had increased to 2 percent. By 1992-1993, total fertility had reportedly fallen to 6.0, desired family size had fallen to 5.9, and modern contraceptive prevalence had increased to nearly 5 percent. Thus it is only in the most recent period that desired family size in Senegal has fallen below actual fertility, and that use of modern contraception has increased above trivial levels.

Changes in desired family size in Senegal have been particularly rapid for women with primary education (falling from about 7.5 children in 1978 to about 4.9 children in 1992-1993), though women with no education show sharp declines as well (from 8.5 to 6.4). Desired family size for women with secondary or higher education was already relatively low in 1978, at 5.0, and declined only to 4.1 by 1992-1993. In Tunisia in 1978, desired family size for women with no education was as low as 4.4, declining slightly to 3.8 by 1988; the change for women with secondary or higher education was even smaller, from 3.1 to 2.8. Fertility measures and family-size preferences in 1992-1993 in Senegal are clearly higher than in Tunisia 15 years earlier, so fertility decline is further behind than child mortality decline. However, all educational categories of women in Senegal, except

those with no education at all, now have a desired family size below the current fertility rate. Given that over a quarter of women of reproductive age in Senegal in 1992-1993 had some education, there is a potential for a substantial decline in fertility within marriage over the next 5 years.

Such fertility decline as has occurred in Senegal appears to have resulted from rising age at first marriage, rather than from control of fertility within marriage. This pattern bears some resemblance to the fertility decline in northern Africa in the 1960s and 1970s. Comparisons of Senegal and Tunisia indicate that the pace of child mortality decline has been very similar in the two countries, with Senegal about 15 years behind. In terms of fertility variables—age at first birth, contraceptive use, and family-size desires—Senegal seems to be somewhat more than 15 years behind Tunisia. However, sharply declining family-size preferences among women with primary education or less over the period 1978 to 1992-1993 suggest that fertility change through contraceptive use may occur in the near future in Senegal.