explain this pattern, gynecological disorders such as pelvic inflammatory disease and the presence of sexually transmitted diseases have been implicated. The other countries with lower than expected fertility rates in 1980 were Egypt, Tunisia, South Africa, and Lesotho. Egypt began to experience a sustained fertility transition as early as 1960 (Omran 1973). Low fertility in South Africa can be explained by its high socioeconomic development.

The Fertility Transition in Africa Author(s): Ezekiel Kalipeni Source: Geographical Review, Vol. 85, No. 3 (Jul., 1995), pp. 286-300 Published by: Taylor & Francis, Ltd. Stable URL: https://www.jstor.org/stable/215274 Accessed: 06-04-2020 22:24 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide

range of content in a trusted digital archive. We use information technology and tools to increase productivity and

facilitate new forms of scholarship. For more information about JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at

Taylor & Francis, Ltd. is collaborating with JSTOR to digitize, preserve and extend access to Geographical Review

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

THE FERTILITY TRANSITION IN AFRICA*

EZEKIEL KALIPENI

ABSTRACT. Some African countries may be going through the initial stages of the fertility transition. In this article multivariate analysis based on country-level data from 1980 and 1993 assesses spatial variations and changes in fertility rates. Demographic and socioeconomic factors such as education, rural or urban resi- dence, status of women, and use of contraceptives are important factors in determining the onset of the fertility transition. Over the long term, fertility will decline to acceptable levels as Africa continues to experience socioeconomic and cultural changes. Of special importance in the transition is the status of women in society. Key words: Africa, contraceptive prevalence, female autonomy, fertility transition, infant mortality rates, population growth.

D uring the past twenty years apparent high growth rates and their implications for the environment have been important emphases

in population studies. A growing popular consensus assumes that high population growth rates in Africa are adversely affecting the environ- ment (Mott and Mott 1980; Goliber 1989; Jolly 1994; Kalipeni 1994; Shapiro 1995). According to this consensus, the increasing pressure of population on limited resources results in destruction of the carrying capacity and hence in declining standards of living. Other interpretations contend that in an efficient market a growing population can encourage innovation and the development of advanced technologies (Boserup 1981; Simon 1983; Shipton 1989).

These debates have determined how scholars study the dynamics of fertility in Africa and have influenced policies about fertility. According to the consensus that high fertility rates and rapid population growth are the major factors in environmental degradation and a declining quality of life, investment of large financial resources in family-planning activi- ties is justified. The study of fertility patterns has emphasized why the rates remain high despite a substantial decline in both adult and infant mortality rates since the 1950s (Okore 1987; Omideyi 1987; Udjo 1987; Mhloyi 1987).

This article adds the largely neglected spatial dimension to the discus- sion of African fertility through a geographical or spatial-temporal frame- work for the precepts of the demographic transition theory. Data from demographic and health surveys (DHS) indicate that fertility levels in Africa have begun to decline, and research supports the thesis that some

* I thank Ellen Kraly and Eliya Zulu for insightful comments on a draft of this article. I am very grateful to Zhen Hou for her untiring research efforts. A grant from the University of Illinois Research Board funded the research.

DR. KALIPENI is an assistant professor of geography at the University of Illinois, Urbana, Illinois 61801.

Copyright ? 1996 by the American Geographical Society of New York

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

FERTILITY TRANSITION IN AFRICA

African countries-Zimbabwe, Botswana, Kenya, and Nigeria, for exam- ple-may be in the initial stages of an irreversible fertility transition (van de Walle and Foster 1990; Cross, Obungu, and Kizito 1991; Caldwell, Orubuloye, and Caldwell 1992).

The central objective of this article is to examine the geographical variation in fertility rates by country throughout Africa in order to iden- tify some of the underlying influences. Quantitative techniques are used to provide confidence in any inferences drawn from the analyzed data. Cultural and socioeconomic factors are hypothesized to be important spatial correlates in the variation of fertility and consequently the onset of the fertility transition now under way.

DATA AND METHODS OF ANALYSIS

Multivariate analysis explains the observed patterns of African fertil- ity. Country-level data from 1980 and 1993 are used in this analysis to assess the spatial variation and change in fertility rates. Socioeconomic and demographic factors include education, rural or urban residence, income levels, status of women as measured by the HDI, infant mortality rates, and contraceptive prevalence. Pairwise t-test for means, analysis of variance, correlation analysis, and stepwise multiple-regression tech- niques are employed to determine the levels of regional variations and to account for the observed spatial variation of fertility rates. The continent is divided into five regions-north, west, east, middle, and south-as suggested by the Population Reference Bureau. These regions exhibit different sociocultural and economic characteristics.

Initially, the spatial relationships among fourteen independent vari- ables and six measures of fertility were examined (Table I). The choice of variables was dictated by the availability of data; therefore, the inde- pendent variables may not necessarily be the best. Also, the quality of data depends on their source. However, data from other sources, espe- cially the recently completed demographic and health surveys, show similar trends and corroborate the findings of this study.

The choice of variables was also based on the causal model of demo-

graphic transition. Proponents of this model generally agree that socio- economic development is the basic cause of fertility decline (Beaver 1975). The model highlights at least four phenomena that are interrelated via social, economic, and psychological mechanisms: urbanization, educa- tion, nonkinship institutions, and consumption levels or standards of living. Social structure includes the relaxation of gender-role restrictions on women, the decreased predominance of extended kinship systems, and the reduced value of children. Furthermore, the model accounts for the effect of other demographic variables such as infant mortality and technological advances on contraceptive availability and use. To opera- tionalize this model, six multivariate linear models stressing the nature

287

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

THE GEOGRAPHICAL REVIEW

TABLE I-DEPENDENT AND INDEPENDENT VARIABLES

Dependent variables Crude birthrate 1980 Crude birthrate 1993

Percent change in crude birthrate 1980-1993 Total fertility rate 1980 Total fertility rate 1993 Percent change in total fertility rate 1980-1993 Pairwise difference in crude birthrate 1980-1993

Pairwise difference in total fertility rate 1980-1993

Independent variables Socioeconomic

Gross national product 1980 Gross national product 1993 Primary-school enrollment ratio 1980 Percent urban population 1980 Percent urban population 1993 Percent change in urban population 1980-1993 Human development index for females

Demographic Infant mortality rate 1980 Infant mortality rate 1993 Percent change in infant mortality 1980-1993 Overall program effort score and family planning Policy concerning fertility reduction Prevalence of modem contraceptives Governmental view of fertility levels

and direction of the relationship between the chosen demographic and socioeconomic variables were formulated. Their form is:

Y = a + PIX1 + -2X2 + p3X3 … E

where Y is the dependent variable, a measure of fertility; a is the Y-inter- cept or constant; p is the regression coefficient; X is the independent variable associated with fertility; and e is the error term.

A FACILE DEMOGRAPHIC TRANSITION?

A World Bank study in 1994 noted the dramatic improvement of health in sub-Saharan Africa during the previous two decades (Shaw 1994). The infant mortality rate had declined by 33 percent, from a high of 145 infant deaths per 1,000 live births in 1970 to 104 per 1,000 live births in 1992. The mortality rate for children under the age of five fell 17 percent between 1975 and 1990. In low-income Africa, between the late 1970s and the late 1980s mortality for the under-five age cohort declined 41 percent in Botswana, 32 percent in Burundi, 31 percent in Mali, and 33 percent in Senegal. Declines have been more rapid in North Africa, with mortality of the under-five age cohort dropping by 50 percent in Egypt and by

288

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

FERTILITY TRANSITION IN AFRICA

around 40 percent in Morocco and Tunisia over the same period (Sullivan 1991). Life expectancy at birth is currently estimated to be 52 years for men and 55 years for women and has increased approximately 4 years every decade since the 1950s (African Development Bank 1992; Popula- tion Reference Bureau 1993). These gains are remarkable, but the chal- lenges of providing accessible health care are immense. The risk of death remains markedly higher at all ages in Africa than it does in other major world regions.

It has been contended that sub-Saharan Africa has not yet experienced a genuine demographic transition and that it is doubtful whether the continent will achieve the transition in a timely fashion (Teitelbaum 1975). The main factor in achieving the transition is socioeconomic develop- ment. The demographic transition postulates a necessary, causal link between modernization and fertility reduction. It explains fertility and population growth solely in socioeconomic terms: the consequences of widespread preference for fewer children that is consequent to industri- alization, urbanization, increased literacy, and declining infant mortality.

Although mortality has declined rapidly in sub-Saharan Africa over the past fifty years, the declines have occurred not because of socioeconomic development but mainly because of the importation of medical technolo- gies from the industrialized world. It therefore can be argued that the declining infant mortality rates and crude death rates on the African continent are due largely to superficial demographic and epidemiologi- cal social changes. Sustainable socioeconomic development has yet to take root on the continent. In an examination of the available data on the

main causes of death, infectious, parasitic, diarrheal, respiratory, and nutritional diseases are prominent, an indication that Africa is still in the age of famine and pestilence, as postulated by the epidemiological transition. In any population the main causes of death are related to the levels of economic and institutional development. The changes in health problems that come with economic and social advancement or a shift in the most common causes of death in a society as it accumulates wealth are often called the mortality or epidemiological transition. Pessimists contend that deep-rooted cultural forces may prevent Africa from ever achieving the fertility transition. Counterarguments can also be offered, especially when one takes into consideration that the demographic transition does not give a time framework and that European countries took more than a century and a half to go through the various stages of the transition. Mortality declines in Africa began only about fifty years ago, so, given more time, the transition may not be out of reach.

FERTILITY LEVELS AND REGIONAL VARIATIONS

By world standards fertility levels on the African continent are still very high. If one excludes the islands of Reunion, Seychelles, Sao Tome,

289

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

THE GEOGRAPHICAL REVIEW

and Principe, total fertility rates in 1993 ranged from 4 to 7 children per woman for most countries. Crude birthrates are also still very high, ranging from a low of 25 births per 1,000 persons in Tunisia to a high of 52 births per 1,000 persons in Niger (Fig. 1). In 1980 only eight African countries had crude birthrates of between 33 and 43 births per 1,000 persons. All other countries on the continent had crude birthrates of 43 or higher. Of the eight countries with crude birthrates of less than 43, four were in the well-documented infertility belt in central Africa (Doenges and Newman 1989). Although no single overriding factor has been iso- lated to explain this pattern, gynecological disorders such as pelvic inflammatory disease and the presence of sexually transmitted diseases have been implicated. The other countries with lower than expected fertility rates in 1980 were Egypt, Tunisia, South Africa, and Lesotho. Egypt began to experience a sustained fertility transition as early as 1960 (Omran 1973). Low fertility in South Africa can be explained by its high socioeconomic development.

By 1993 fertility had declined considerably in most parts of Africa, with regional patterning in the changes. Northern and southern Africa stand out as regions that have experienced the greatest declines in fertility rates. On the other hand, countries in central Africa experienced slight increases in their fertility rates, which is also true for Sahelian countries of western Africa. Analysis of variance for the difference between re- gional means yields a regional patterning for crude birthrates statistically significant at the 5 percent level in 1993 and a regional difference in means for total fertility rates significant at the 10 percent level for the same year (Table II). Paired comparisons of t-tests for the change in fertility rates between 1980 and 1993 indicate statistically significant declines at the continental level, especially for northern and southern Africa (Table III). The eastern and western regions experienced declines that were statisti- cally insignificant. On the other hand, central Africa experienced a slight but insignificant increase in crude birthrates. A similar analysis of total fertility rates yields more or less the same results (Table IV).

At the country level, fertility declines of varied magnitude affected most countries (Table V and Fig. 2). Data from demographic and health surveys and other published works indicate that fertility levels in most African countries have indeed begun to decline. For example, in Bo- tswana the total fertility rate declined from 7 to 5, or 30 percent, between 1981 and 1988 (Lesetedi, Mompati, and Khulumani 1989). Kenya, which has long experienced the highest population growth rate in Africa-more than 4 percent annually-appears to be on the way to a fertility transition (Kelley and Nobbe 1990). By 1989 the fertility rate for Kenya had declined by 17 percent to a low of 6.7 (Kenya Ministry of Home Affairs and National Heritage 1989; Cross, Obungu, and Kizito 1991). Between 1984 and 1988 total fertility rates in Zimbabwe had declined from 6.5 to 5.5

290

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

FERTILITY TRANSITION IN AFRICA 291

Computation CBR, 1the author;data from Population Refere980nCBR 19931993

3 3300 to43. 085 40.90 to 44.7 10

43.00 to 47.00

47.00 to 49.00

PercFIG. Crude birrates for 1980 and 1993 and percentage change in crude birrates. Sources:

(Meekers 1991). In addition, the fertility transition may also have started in some areas of other countries, including Nigeria, Ghana, Senegal, Malawi, and Tanzania (Ghana Statistical Service 1989; Ngallaba and others 1993; Kalipeni and Harrington 1995). The DHS survey data also

0.67 to 4.7.

4.78O0o 23.65

indicate that feercentility appeahave increased in some countries.

Computation by the author; data from Population Reference Bureau 1980, 1993.

indicate that fertility appears to have increased in some countries. In Namibia the total fertility rate is 5, and fertility in this country has

been declining gradually over the past fifteen years (Katjiuanjo and

.s

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

THE GEOGRAPHICAL REVIEW

TABLE II-ANALYSIS OF VARIANCE FOR FERTILITY RATES IN AFRICA, BY REGION, 1980 AND 1993

MEAN CRUDE MEAN CRUDE MEAN TOTAL MEAN TOTAL REGION BIRTHRATE 1980 BIRTHRATE 1993 FERTILITY RATE 1980 FERTILITY RATE 1993

North 42 34 6.2 4.88 West 49 46 6.8 6.71 East 48 47 6.6 6.77 Middle 45 46 6.0 6.46 South 39 35 5.2 4.57

F-ratio 1.13 4.61 0.72 2.45 P-value 0.3529 0.0032a 0.5825 0.0590b

Sources: Computation by the author; data from Population Reference Bureau 1980, 1993. a Statistically significant at the .05 level. b Statistically significant at the .10 level.

TABLE III-PAIRED COMPARISONS T-TEST FOR DIFFERENCES IN CRUDE BIRTHRATE IN AFRICA, BY REGION, 1980 AND 1993

MEAN DIFFERENCE IN

REGION CRUDE BIRTHRATE STANDARD ERROR T-STATISTIC P-VALUE

North -9.12 1.62 -5.62a 0.0049 West -0.870 .95 -0.91 0.3767

East -1.36 1.64 -0.83 0.4265 Middle 1.38 1.16 1.19 0.2736 South -4.07 1.38 -2.96a 0.0161 Africa -2.04 0.70 -2.92a 0.0052

Sources: Computation by the author; data from Population Reference Bureau 1980, 1993. a T-statistic is significant at the .05 level.

TABLE IV-PAIRED COMPARISONS T-TEST FOR DIFFERENCES IN TOTAL FERTILITY RATE IN AFRICA, BY REGION, 1980 AND 1993

MEAN DIFFERENCE IN

REGION TOTAL FERTILITY RATE STANDARD ERROR T-STATISTIC P-VALUE

North -1.82 0.40 -4.50a 0.0108

West 0.06 0.14 0.39 0.7013

East -1.36 0.36 -0.25 0.8056

Middle 0.25 0.17 1.50 0.1778

South -0.59 0.20 -2.89a 0.0179 Africa -0.25 0.13 -1.89 0.0646

Sources: Computation by the author; data from Population Reference Bureau 1980, 1993. a T-statistic is significant at the .05 level.

others 1993). In Malawi the total fertility rates declined from 8 in 1987 to 7 in 1992, for a decrease of 12 percent in just five years (Malawi National Statistical Office 1992). In Zambia total fertility rates declined from a high of 7 in the 1980 census to 6.5 in 1992 (Gaisie, Cross, and Nsemukila 1993).

292

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

FERTILITY TRANSITION IN AFRICA

TABLE V-TOP 15 AFRICAN COUNTRIES WITH GREATEST DECUNES IN CRUDE BIRTHRATES FROM 1980 TO 1993

PERCENT

PERCENT CHANGE IN PERCENT

CHANGE IN INFANT PERCENT CHANGE IN MODERN

CRUDE CRUDE CRUDE MORTALITY CHANGE IN URBAN CONTRACEPTIVE BIRTHRATE BIRTHRATE BIRTHRATE RATE GNP POPULATION PREVALENCE

COUNTRY 1980 1993 1980-1993 1980-1993 1980-1993 1980-1993 IN 1993

Algeria 48.0 34.0 -29.17 -57.04 60.31 -9.82 31 Botswana 51.0 36.9 -27.65 -54.02 317.74 110.83 32 Morocco 43.0 31.3 -27.21 -56.84 53.73 10.71 36 Mauritius 27.0 20.7 -23.33 -46.86 191.57 -6.82 46

Tunisia 33.0 25.4 -23.03 -65.60 58.95 18.40 40

Sao Tome & Principe 45.0 35.0 -22.22 -12.34 -28.57 76.25 Rwanda 50.0 39.5 -21.00 -13.46 44.44 35.00 13

Egypt 38.0 30.8 -18.95 -37.78 55.00 -0.23 44 Kenya 53.0 45.2 -14.72 -12.89 6.25 136.00 18 Malawi 51.0 44.0 -13.73 -21.42 27.78 74.72 7

Zimbabwe 47.0 40.6 -13.62 -54.26 29.16 30.14 36

Lesotho 40.0 35.4 -11.50 -24.32 107.14 372.50

Libya 47.0 41.9 -10.85 -47.69 -15.30 26.33 Ghana 48.0 43.0 -10.42 -25.65 2.56 -10.00 5

Nigeria 50.0 44.8 -10.40 -46.31 -48.21 -19.50 4

Sources: Computation by the author; data from Population Reference Bureau 1980, 1993.

Even in the predominantly Muslim, northern African countries of Morocco, Tunisia, and Egypt, fertility declined by 18 percent or more. Libya, which has been noted for its persistent high fertility rate, experi- enced a modest decline of 10 percent. These declines may not be dramatic, but they indicate that unusually high fertility rates in Africa are not static.

REGRESSION ANALYSIS

To account for the observed geographical variations in fertility levels on the continent, six stepwise regression models were generated with the dependent and independent variables (Table VI). As indicated by the F-ratio and the associated R-square values, all six models were statisti- cally significant at the 5 percent level. The first model concerned the variation of total fertility rates in 1980. Only one independent variable met the criterion for entry into a stepwise regression method. GNP per capita for 1980, the variable chosen, explained a modest 15 percent of the variation in total fertility rates for that year. The second model used total fertility rates in 1993 as the dependent variable. Three independent variables-total fertility rates in 1980 included as a control variable, the HDI for females in 1990, and modern contraceptive prevalence in 1993- were selected for entry into the model by the stepwise procedure in the SAS statistical software package. All standardized regression coefficients of the three independent variables were statistically significant at the 5 percent level. The signs of the standardized regression coefficients are

293

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

THE GEOGRAPHICAL REVIEW

Mali –

-C– —— — ^ – i TFR 1990 Uganda . .1 Malawi j TFR 1980 i- _ .

Burundi

Kenya

Liberia

TaZambia

– I

Z Senegal

Togo

Ghana

Tanzania

Nigeria

Zimbabwe

Namibia

Botswana

1 24 5 6 8

TOTAL FERTILITY RATES (TFR)

FIG. 2-Total fertility rates for selected DHS countries 1980 and 1990. Sources: Computation by the author; data for 1980 from Population Reference Bureau 1980, and for 1990 from various DHS publications.

also in the hypothesized direction. In other words, the greater the HDI, the smaller the total fertility rate, a negative relationship; the greater the percentage of population using modem contraceptives, the smaller the fertility rate, a negative relationship; and the greater the total fertility rate in 1980, the larger the fertility rate in 1993, a positive relationship. This model explained 71 percent of the variation in fertility rates for 1993.

The third and fourth models used crude birthrates in 1980 and 1993 as

independent variables. The results were similar to those obtained for total fertility rates. GNP per capita in 1980 and percentage of urban population accounted for 29 percent of the variation in crude birthrates in 1980, with the standardized coefficient for GNP statistically significant at the 5 percent level. Both variables chosen for entry into the model for crude birthrates in 1993 had statistically significant regression coefficients. Once again the HDI and the prevalence of modern contraceptives proved to be powerful explanatory variables for crude birthrates in 1993.

The fifth and sixth models used the percentage change in fertility rates between 1980 and 1993 as the dependent variable. The fifth model shows that 51 percent of the change in crude birthrates between 1980 and 1993

294

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

FERTILITY TRANSITION IN AFRICA

TABLE VI-STANDARDIZED PARTIAL REGRESSION COEFFICIENTS*

DEPENDENT VARIABLES

Change in Change in Total

Crude Fertility Total Total Crude Crude Birthrate Rate

Fertility Fertility Birthrate Birthrate 1980-1993 1980-1993 INDEPENDENT VARIABLE Rate 1980 Rate 1993 1980 1993 (%) (%)

GNP per Capita 1980 -0.3857a -0.3993a 0.3129a Total Fertility Rates 1980 0.3957a Percent Urban

Population 1980 -0.3993 -0.2080 -0.2267 Infant Mortality Rate 1980 0.5949a 0.5188a Human Development Index for Females 1990 -0.2868a -0.3326a

Modem Contraceptive Prevalence 1993 -0.5582a -0.5620a

Change in Infant Mortality Rates 1980-1993 (%) 0.3434a 0.2603a

r2 0.15 0.71 0.29 0.64 0.51 0.41 F-ratio 6.29a 30.9a 7.27a 34.8a 8.40a 7.87a N 37 41 41 41 37 37

* Empty cells are variables not selected for entry into model by stepwise procedure. a Statistically significant at the .05 level.

can be explained by four independent variables-GNP in 1980, percent- age urban population in 1980, level of infant mortality rate in 1980, and change in infant mortality rates between 1980 and 1993. In the sixth model, 41 percent of the change in total fertility rates between 1980 and 1993 was accounted for by three variables: percentage urban population, infant mortality rates, and change in infant mortality rates. For all of the models in which percentage urban population was entered, the stan- dardized coefficient was not significant at the 5 percent level but was so at the 10 percent level.

POLICY IMPLICATIONS

This overview of regression results highlights the importance of both socioeconomic and demographic variables in fertility reduction. Im- provements in social, economic, and demographic conditions are cru- cial to the realization of an irreversible, sustainable fertility transition in Africa. Clearly, contraceptive prevalence and the status of women as measured by the HDI as well as reductions in infant mortality rates are central to the ongoing fertility declines in many parts of Africa. My discussion concentrates on these three variables and their implications for a fertility transition on the continent.

295

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

THE GEOGRAPHICAL REVIEW

THE HUMAN DEVELOPMENT INDEX AND STATUS OF WOMEN

The United Nations Development Programme formulated the HDI as a composite measure of economic and social welfare. Unlike other meas- ures of socioeconomic development, this index gives equal weight to longevity, educational attainment, and utility derived from income (UNDP 1990). A score is derived for each of these, from 0 for the lowest achieved by any country to 1 for the highest. Averaging the three indica- tors gives a HDI value between 0 and 1. The Human Development Report 1990 (UNDP 1990) gives the index by sex for each country in Africa, and by extracting the data I compiled a HDI for females as a surrogate measure of the status of women across the continent. In the regression analysis the HDI was a strong explanatory variable for the spatial pat- terns of fertility in 1993 as measured by total fertility rates and crude birthrates. Countries that had a high HDI exhibited a lower fertility rate.

The importance of status and autonomy of women in the attainment of fertility reductions cannot be overemphasized. If a woman is the main controller of her reproductivity, she is more likely to use contraceptives and to limit the ultimate number of children born than if the decision is

left to a man, because women assume most of the physiological and child-care burdens of frequent childbearing. One reason for the persist- ence of high fertility in sub-Saharan Africa is the minimal involvement of women in decision making about childbearing.

The main policy implication of this finding is to reiterate the call for upgrading the status of African women. This variable decisively shows that in countries in which the status of women has improved, declines in fertility have been dramatic during the past decade. Fertility decreases as the education of females increases.

INFANT MORTALITY

A good indicator of health conditions in a country is its infant mortal- ity rate: the number of deaths of children under the age of one per 1,000 live births annually. Studies during the past twenty years or so indicate that a very significant relationship exists between high levels of infant and child mortality and low levels of maternal education (Caldwell 1979; Cleland and Van Ginneken 1989; Bicego and Fegan 1991; Kalipeni 1993). The education of females seems to correlate highly with infant mortality rates. Even if incomes are low, educated women tend to make better decisions about their children than do uneducated women.

Lower levels of infant mortality have been shown to be strongly correlated with lower fertility rates. Because of comparatively high child- hood mortality throughout the developing world, many families have felt the need to have more children to ensure that some will survive to adulthood. This pattern is further reinforced by the need for sons as social security in old age.

296

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

FERTILITY TRANSITION IN AFRICA

The importance of infant mortality in any fertility transition is high- lighted by the situation in Kenya, Zimbabwe, and Botswana, each of which has an infant mortality rate below 70 per 1,000 live births (Caldwell, Orubuloye, and Caldwell 1992). No other countries in sub- Saharan Africa record a level below 80, and most have rates above 100, which suggests that the attainment of this level of infant mortality may prove to be the necessary condition for African fertility decline. In this analysis, infant mortality rates and change in infant mortality rates had a significant effect on total fertility rates. Countries with the greatest declines in infant mortality rates between 1980 and 1993 also experienced the greatest fertility declines. This pattern again can easily be related to female autonomy and status.

The policy option is clear: the role and status of women should be changed through vigorous pursuit of universal education of girls through- out the continent. Governments are urged to increase women’s access to education, employment opportunities, and extension and credit services and to enact laws that protect women’s rights against male domination in various sectors of economic and social life.

CONTRACEPTIVE PREVALENCE AND FAMILY PLANNING

Perhaps of greatest significance in the onset of the fertility transition in the top fifteen countries in decline of crude birthrates is the use of modern and effective contraceptives. During the 1960s and 1970s African governments were reluctant to institute effective family-planning pro- grams, and political support for them in the public sector was “uniformly weak” throughout the continent (Africa Development Bank 1992). This attitude has changed, however; and since the 1974 and 1984 world population conferences, governments in several African countries have acknowledged that they consider fertility levels to be high. Most govern- ments with this viewpoint have initiated family-planning programs. Although success in popularizing the use of modern contraceptive tech- niques has been slow in coming, some countries have built an impressive infrastructure for provision of these services.

In 1973 the government of Botswana launched a maternal- and child- health program into which family-planning services were integrated. This program has been very successful in achieving its goals and objec- tives. By 1988, for 92 percent of births, mothers had received prenatal care; for 66 percent, mothers had medically supervised deliveries; and for 54 percent, mothers had received postnatal care (Lesetedi, Mompati, and Khulumani 1989). From 1984 to 1988 current use of modem methods of contraception among married women aged fifteen to forty-nine rose from 19 to 29 percent, and total fertility rates declined from 7 to 5. Data from Zimbabwe show that the decline of fertility there has been caused by changes in both marriage patterns and contraceptive use. The shift to use

297

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

THE GEOGRAPHICAL REVIEW

of modern contraception practices seemed most pronounced among un- educated rural populations, which implies that neither education nor other forms of socioeconomic change were responsible. This pattern indicates that if a fertility transition is to be achieved, there will be a great need for the provision of effective methods of contraception in both rural and urban settings.

Investigations in Kenya found large variations in contraceptive use among the thirteen districts covered in the DHS data (Kelley and Nobbe 1990; Cross, Obungu, and Kizito 1991). These findings indicate that pockets of contraceptive use in Kenya, Zimbabwe, and Botswana reach levels as high as those found in developed countries and that these areas are rapidly going through a fertility transition (Way, Cross, and Kumar 1987). Throughout the continent there are signs of progress in contracep- tive use. The celebrated family-planning programs in Botswana, Zim- babwe, South Africa, and Kenya have followed different paths, but all have succeeded. In Botswana, for example, family-planning services operate in the framework of an integrated maternal- and child-health program of a decentralized health system. The Zimbabwe National Fam- ily Planning Council is a semiautonomous unit in the Ministry of Health (Lucas 1992). The common theme in the southern African programs is a commitment to the provision of safe and effective means of child spacing or of stopping births altogether.

In short, successful programs, coupled with strong political support for family planning, have undermined the preference for large families. Rapid declines in infant mortality rates, improved maternal and child care, rapidly increasing attainment of formal education by females, and increased use of effective contraceptives have all contributed to the onset of a fertility transition, especially in northern and southern Africa.

CONCLUSION

Through analysis of variance, paired t-tests, and exploratory stepwise regression analysis this article demonstrates that northern and southern Africa are in the process of a fertility transition. Furthermore, the results demonstrate that a significant number of countries on the continent experienced substantial declines in their fertility rates between 1980 and 1993. Female autonomy as measured by the HDI, total fertility rates, percentage of urban population, infant mortality rates, and moder con- traceptive prevalence are strongly associated with the declining fertility rates. Countries that scored very high on the HDI also experienced the greatest declines in fertility rates and had relatively lower fertility rates. Countries that experienced the greatest declines in infant mortality rates also exhibited the greatest declines in fertility rates. Although fertility rates are still very high by world standards and population will continue to grow at high rates, the declining trends need to be reinforced through

298

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

FERTILITY TRANSITION IN AFRICA

the education of women, the reduction of high infant mortality rates, and the provision of safe and effective contraception.

This analysis refutes the argument that African countries may not achieve a fertility transition in the near future, because the necessary socioeconomic conditions for the transition are weak or nonexistent. The

achievement of the fertility transition is possible if certain policy issues are addressed with unyielding governmental commitment. Over the long term, fertility will decline to acceptable levels as Africa continues to experience socioeconomic and cultural transformation. Of great impor- tance in the timely achievement of the transition is the status of women in society.

CITATIONS

African Development Bank. 1992. African development report 1992. Development Research and Policy Department, Abidjan, Ivory Coast.

Beaver, S. E. 1975. Demographic transition theory reinterpreted. Lexington, Mass.: Lexington Books.

Bicego, G. T., and C. Fegan. 1991. Maternal education and child survival: a comparative analysis of DHS data. Demographic and health surveys world conference: proceedings, vol. 1, 177-204. Columbia, Md.: IRD/Macro International.

Boserup, E. 1981. Population and technological change. Chicago: University of Chicago Press. Caldwell, J. C. 1979. Education as a factor in mortality decline: an examination of Nigerian data.

Population Studies 33:395-413. Caldwell, J. C., I. 0. Orubuloye, and P. Caldwell. 1992. Fertility decline in Africa: a new type of

transition. Population and Development Review 18:211-242. Cleland, J. G., and J. K. Van Ginneken. 1989. Maternal schooling and childhood mortality. Journal

of Biosocial Science supplement 10:13-34. Cross, A. R., W. Obungu, and P. Kizito. 1991. Evidence of a transition to lower fertility in Kenya.

International Family Planning Perspectives 17:4-7. Doenges, C. E., and J. L. Newman. 1989. Impaired fertility in tropical Africa. Geographical Review

79:99-111.

Gaisie, K., A. R. Cross, and G. Nsemukila. 1993. Zambia demographic and health survey 1992. Lusaka: University of Zambia and Zambia Central Statistical Office.

Ghana Statistical Service. 1989. Ghana demographic and health survey 1988. Accra. Goliber, T. J. 1989. Africa’s expanding population: old problems, new policies. Population Bulletin

44(3):1-51. Jolly, C. L. 1994. Four theories of population change and the environment. Population and Environ-

ment 16:61-91.

Kalipeni, E. 1993. Determinants of infant mortality in Malawi: a spatial perspective. Social Science and Medicine 37:183-198.

Kalipeni, E., ed. 1994. Population growth and environmental degradation in southern Africa. Boulder, Colo.: Lynne Rienner.

Kalipeni, E., and L. Harrington. 1995. Regional variations of fertility in Malawi. Scandinavian Journal of Development Alternatives 14(1/2):222-247.

Katjiuanjo, P., S. Titus, M. Zauna, and J. T. Boerma. 1993. Namibia demographic and health survey 1992. Windhoek: Namibia Ministry of Health and Social Services.

Kelley, A., and C. Nobbe. 1990. Kenya at the demographic turning point? World Bank discussion paper no. 107, World Bank, Washington, D.C.

Kenya Ministry of Home Affairs and National Heritage. 1989. Kenya demographic and health survey, summary report. Nairobi.

Lesetedi, L., G. Mompati, and P. Khulumani. 1989. Botswana family health survey II, 1988. Ga- borone: Central Statistical Office, Ministry of Finance and Development Planning, and Ministry of Health.

Lucas, D. 1992. Fertility and family planning in southern Africa and central Africa. Studies in Family Planning 23(3):145-158.

Malawi National Statistical Office. 1992. Malawi demographic and health survey 1992. Zomba.

299

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

THE GEOGRAPHICAL REVIEW

Meekers, D. 1991. Report on the IUSSP seminar on the course of fertility transition in sub-Saharan Africa, Harare, Zimbabwe, 19-22 November 1991. IUSSP paper no. 31, International Union for the Scientific Study of Population, Liege, Belgium.

Mhloyi, M. 1987. The proximate determinants and their socio-cultural determinants: the case of two rural settings in Zimbabwe. The cultural roots of African regimes: proceedings of the Ife conference, February 25-March 1, 1987. Ife, Nigeria: Obafemi Awolowo University, Department of Demography and Social Statistics.

Mott, F. L., and S. H. Mott. 1980. Kenya’s record population growth: a dilemma of development. Population Bulletin 35(3):1-43.

Ngallaba, S., S. H. Kapiga, I. Ruyobya, and J. T. Boerma. 1993. Tanzania demographic and health survey 1991/1992. Dar-es-Salaam: Tanzania Bureau of Statistics Planning Commission.

Okore, A. 0. 1987. Fertility determinants among the Igbo in eastern Nigeria. The cultural roots of African regimes: proceedings of the Ife conference, February 25-March 1, 1987. Ife, Nigeria: Obafemi Awolowo University, Department of Demography and Social Statistics.

Omideyi, A. K. 1987. Status, cultural beliefs and fertility behavior among Yoruba women. The cultural roots of African regimes: proceedings of the Ife conference, February 25-March 1, 1987. Ife, Nigeria: Obafemi Awolowo University, Department of Demography and Social Statistics.

Omran, A. R. 1973. Egypt’s fertility profile. Egypt: population problems and prospects, ed. A. R. Omran, 73-114. Chapel Hill: University of North Carolina, Carolina Population Center.

Population Reference Bureau. 1980. World population data sheet. Washington, D.C. . 1993. World population data sheet. Washington, D.C.

Shapiro, D. 1995. Population growth, changing agricultural practices, and environmental degrada- tion in Zaire. Population and Environment 16(3):221-236.

Shaw, P. 1994. Better health in Africa: experiences and lessons learned. Washington, D.C.: World Bank.

Shipton, P. 1989. How private property emerges in Africa: directed and undirected land tenure reforms in densely settled areas south of the Sahara. Report to the Bureau of Program and Policy Coordination, USAID. Cambridge, Mass.: Harvard University.

Simon, J. 1983. Population pressure on the land: analysis of trends past and future. World Develop- ment 11(9):825-834.

Sullivan, J. M. 1991. The pace of decline in under-five mortality: evidence from the DHS surveys. Demographic and health surveys world conference: proceedings, vol. 1, 25-38. Columbia, Md.: IRD/Macro International.

Teitelbaum, M. S. 1975. Relevance of the demographic transition theory for developing countries. Science 188:420-425.

Udjo, E. 0. 1987. A determinant of fertility and its cultural context among Nigeria’s Kanuri. The cultural roots of African regimes: proceedings of the Ife conference, February 25-March 1, 1987. Ife, Nigeria: Obafemi Awolowo University, Department of Demography and Social Statistics.

UNDP (United Nations Development Programme). 1990. Human development report. New York: Oxford University Press.

van de Walle, E., and A. D. Foster. 1990. Fertility decline in Africa: assessment and prospects, World Bank technical paper no. 125, Africa Technical Department Series, World Bank, Washington, D.C.

Way, A. A., A. R. Cross, and S. Kumar. 1987. Family planning in Botswana, Kenya, and Zimbabwe. International Family Planning Perspectives 13:7-11.

300

This content downloaded from 107.77.201.20 on Mon, 06 Apr 2020 22:24:36 UTC All use subject to https://about.jstor.org/terms

  • Contents
    • image 1
    • image 2
    • image 3
    • image 4
    • image 5
    • image 6
    • image 7
    • image 8
    • image 9
    • image 10
    • image 11
    • image 12
    • image 13
    • image 14
    • image 15
  • Issue Table of Contents
    • Geographical Review, Vol. 85, No. 3, Jul., 1995
      • Front Matter
      • Political Stability and Minority Groups in Burma [pp.269-285]
      • The Fertility Transition in Africa [pp.286-300]
      • Indigenous Development of Mesoamerican Irrigation [pp.301-323]
      • Home as a Region [pp.324-334]
      • Mexicali’s Chinatown [pp.335-348]
      • Postmodern Phoenix [pp.349-363]
      • Urban Community Gardens as Contested Space [pp.364-381]
      • Geographical Record
        • Oil and Gas in Austria [pp.382-383]
        • New Pipelines in Australia [pp.384-385]
      • Geographical Reviews
        • untitled [pp.386-388]
        • untitled [pp.388-390]
        • untitled [pp.390-391]
        • untitled [pp.391-392]
        • untitled [pp.392-394]
        • untitled [pp.394-396]
        • untitled [pp.396-398]
        • untitled [pp.398-400]
        • untitled [pp.400-402]
        • untitled [pp.402-403]
        • untitled [pp.403-404]
        • untitled [pp.404-406]
        • untitled [pp.406-408]
        • untitled [pp.408-410]
        • untitled [pp.410-411]
        • untitled [pp.412-413]
      • Back Matter [pp.414-414]