 |
| About Bioline | All Journals | Testimonials | Membership | News |
|
||||||
|
||||||
Union for African Population Studies, Rapport de Synthese / Summary Report, Numéro/Number 7, Mar. 1994 SMALL GRANTS PROGRAMME ON POPULATION AND DEVELOPMENT FAMILY AND REPRODUCTIVE BEHAVIOR IN URBAN TOGO: A SUMMARY Koffi Ekouevi, Ph.D. African Institute for Economic Development and Planning, (IDEP) Dakar, Senegal Sponsored by The UAPS Small Grants Programm on Popula-tion and Development funded by IDRC French Cooperation, Mac Arthur Foundation, Rockfeller Foundation and SAREC Code Number: uaps94007 CONTENTS 1. Introduction 1. Introduction The dominant perspective in studying the African family is based on the assumption that a convergence of family values and behaviors towards characteristics of the Western family will be observed with social change. The exposure to alternative lifestyles through modernization, westernization and urbanization is expected to alter features of the traditional African family. Goode (1963) predicted that the nuclear family will be progressively adopted as modernization occurs. According to Caldwell (1982), westernization is the main social force that will bring family nucleation in Sub-Saharan Africa. He assumed that westernization driven by mass education will push individuals to adopt values and behaviors observed in Western societies. A growing number of studies are questionning the convergence argument (van de Walle and Kekovole, 1984; Boserup, 1985; Locoh, 1988; Lesthaeghe, 1989; Thornton, 1992). Where characteristics of the Western family were expected, the realities of the African family show more complexity. Consequently, Thornton has noted that the expected convergence of family values and behaviors from non-Western societies to those of Western society is the result of our heritage from social scientists of the eighteenth and nineteenth centuries who read family history sideways. He stated: The paradigm that all societies move through very uniform trajectories of social change is very old and has played a significant role in social thought throughout Western history. The paradigm draws its central idea from the metaphor comparing social change to the growth and development of biological organisms. This metaphor was a central feature of ancient Greek thought and was also part of the Roman and early Christian world view, including that of St. Augustine. (Thornton, 1992: 21) The notion of evolutionary or developmental growth of biological organisms was applied to societies. Non-Western societies were classified at the bottom of the ladder while Western societies were placed at the top. From the perspective of these scholars, social patterns of non-Western societies will progressively meet standards of civilization as these societies mature or develop. According to Thornton, scholars of the eighteenth and nineteenth centuries constructed on the basis of these assumptions a powerful theory of societal development that persists in this final decade of the twentieth century. They used information from non-Western societies of their time to represent social life at an early stage of development, and used that of Western societies to represent an advanced stage of development, thereby constructing a universal theory of social change. As a contribution to correcting the errors of past research, Thornton suggests a paradigm shift -- from a developmental paradigm to a non-developmental paradigm. The former holds that as social and economic development occur, the African family will converge to the nuclear model observed in the West. In contrast, the non-developmental paradigm suggests that changes will emerge from traditional lifestyles in response to social and economic conditions and constraints. In essence, Thornton suggests the need to consider plurality or diversity in approaching individual life experiences and societal trajectories. Against this theoretical background this study addresses two fundamental questions: first, what kind of changes are occurring to the family institution in urban Togo ? And second, in what direction are they leading? Specific areas of investigations are marriage and family life (polygyny, spouses' residential arrangements), fertility and postpartum behavior (sexual abstinence, breastfeeding and contraception), and fertility intentions (desire for additional children, ideal family size). This report is a summary of main results from the study. It is organized as follows: at first, the social setting is presented, followed by a description of data and methods. Next, is a summary of main results and finally theoretical and policy implications are outlined. 2. The Social Setting The understanding of behaviors and attitudes regarding family and reproductive behavior in urban Togo requires familiarity with main historical and contextual facts and the process of urbanization in this setting. 2.1 The Country and the People Togo is a small country in West Africa, with a total area of 56,600 square kilometers (21,853 sq. mi.) and a total population of about 3.8 million inhabitants in 1992 (United Nations Population Fund, 1993). It is bordered by Benin on the east, Ghana on the west, Burkina Faso on the north and the Atlantic Ocean on the south. As the majority of the countries in Sub-Saharan Africa, the contemporary territory of Togo results from a complex colonial experience. Between 1884 and 1914 the territory of Togo was part of Togoland, a German colony. At the beginning of the First World War, Germany lost its colony to a joint British and French force. This force occupied Togoland between 1914 and 1919. After the war, the eastern part of Togoland was placed under the control of a French administration and the western part was attributed to the Gold Coast, a British colony by the League of Nations. This political organization remained until 1956. In May 1956 a United Nations-supervised plebiscite voted in support of incorporating the western part of Togoland to the British Gold Coast; this territory became the Republic of Ghana in 1957. In another United Nations plebiscite held in October 1956, the eastern part of Togoland voted for a separate republic that became independent in 1960. A consequence of the arbitrary delineation of boundaries is the diversity of ethnic groups in the country. Despite its small size, Togo regroups about forty different ethnic groups. At the national level, the Aja-Ewe from the South are the dominant group, followed by the Kabye-Tem from the North. At the 1981 Census, the Aja-Ewe represented 44 percent of the total population, and the Kabye-Tem 27 percent (Direction de la Statistique, 1989). All these ethnic groups share the practice of traditional African religion. In this religion, individuals worship God, small divinities, spirits and ancestors also called the living-dead. It is believed that ancestors are spiritually in contact with the living. They remain part of the family and can influence the destiny of the living. Concurrence and protection received from ancestors are believed to be associated with success in individuals' endeavors during their lifetime. Ancestors have a say in many life-cycle events. For example, before marriage ancestors from both sides have to give their agreement through traditional ceremonies. Similarly, before a major trip they have to be consulted to clear the way of bad luck. Also, at the death of a family member, the cause of death must be investigated spiritually so as to identify the evil force that caused the damage and strengthen the protection system. Elderly members of the family are those who have the wisdom to protect the wishes of ancestors and to coordinate communication between the living and ancestors. Besides the traditional African religion, Christianity and Islam are also important forces that influence individuals' lives. However, it can be speculated that the importance of the traditional religion, Christianity, and Islam vary along a spectrum across individuals. Regardless of the adopted religion, Christian or Islamic, traditional beliefs are well rooted in the socio-cultural context in both urban and rural areas. In other words, an individual may be Christian or Muslim but can still perform family rituals originally associated with the traditional religion. 2.2 The Urbanization Process Although cities existed in precolonial Togo, contemporary urbanization, as in most Sub-Saharan African countries, originated as a response to the economic and administrative needs of the metropolis during colonial times. Urban areas were constructed to coordinate the commercial and administrative activities of European colonial countries. The German administration chose Lomé as the capital city of Togoland in 1897. The area was already occupied by groups of Ewe, Ouatchi, and Mina, along with some traders from Kéta and surrounding villages. These coastal inhabitants had been trading with Portuguese, Dutch, Danish, French and British from the 15th century. They also participated in the trade of humans, mainly in the 18th century. As a strategy to organize Togoland, which was considered the Musterkolonie (model colony), the German administration created five Bezirksämtern (administrative circles) with Unterbezirken (administrative subdivisions), and three Stationsbezirke (administrative centers) (Nyassogbo, 1984). Three main railroads were built: the "Coconut line" in 1905, the "Cocoa line" in 1907, and the "Cotton line" in 1911. A wharf was also built in 1900, linking Lomé to the metropolis. Other signs of urbanization in Togoland were the creation roads, the opening of schools, and the construction of hospitals (Nyassogbo, 1984). The British and especially the French built on the German infrastructure after the Germans lost their colony in the First World War. The strategy of urbanization was the same. The city functioned mainly as an administrative and trading center to control the territory and to organize the extraction of raw materials from the hinterlands to support European industries. This remains the character of Lomé some 30 years after independence, since fundamental changes have not been made in the economic and political institutions of the country. As a consequence, the disparity in infrastructure between Lomé and other areas of the country is striking. In the eighties, Lomé was an attractive destination for migrants, a melting pot of all the country's ethnic groups and from some neighboring countries as well. 3. Data and Methods The study is primarily based on data from the APEL (Arrivée du Prochain Enfant à Lomé) survey. The APEL survey was conducted by the URD (Unité de Recherche Démographique) of Togo between June 1983 and February 1984. APEL collected data on demographic, socio-economic, and family background characteristics; birth history data; data on proximate determinants of fertility; and data on fertility preferences. About 3,606 women in their reproductive years were interviewed, regardless of marital status. The data set provides information on these various subjects, but certain questions were asked only of selected groups of women, limiting in-depth investigation on some topics. To support analyses based on the APEL data, data from the EDST (Enquête Démographique et de Santé au Togo) survey is used. The EDST survey was conducted in 1988 as part of the Demographic and Health Surveys program. This survey was also conducted by the Unité de Recherche Démographique of Togo, in collaboration with the Institute for Resource Development in Columbia, Maryland, USA. The objectives of the EDST were to study levels, trends, and differentials in fertility and infant and child mortality; to study determinants of fertility; to provide information on knowledge and use of contraception; and to study reproductive intentions and certain other health indicators. Unlike the APEL data, the EDST data is representative of all women in their reproductive years in the country. About 3,360 women aged 15-49 were interviewed in the whole country, about 750 of them in Lomé. The comparison of data from the two surveys on these selected variables indicates that the APEL data is generally off on levels as compared to the EDST data. However, preliminary results indicate that the APEL data describe fairly well differentials between subgroups. Consequently, in this study, the EDST data is used to discuss levels and trends and to provide a representative picture of Lomé. The APEL data is used mostly to study cross-sectional differences in the eighties. 3.1 Description of Variables Six broad groups of variables are created from the APEL data: 1) Social background 2) Socio-economic status 3) Basic demographic characteristics 4) Family formation and characteristics of unions 5) Proximate determinants of fertility and 6) Fertility intentions. Social background is measured by parental education, childhood residence, and duration of residence in Lomé. These variables are indicators of the socialization environment of women. It is expected that a woman's perception of family life and her reproductive choices are influenced to some extent by her social background. Women coming from a social background with some exposure to modern values are expected to be the ones leading the way toward changes in family and reproductive behavior. For example, women coming from a background where parents are educated are likely to have earlier exposure to modern values that portray the Western nuclear family type as the ideal. Along the same lines, growing up entirely in an urban environment facilitates acquaintance with a variety of family choices in one's lifetime. Unlike the limited boundaries of the village community, the urban area offers a broader life horizon with alternative lifestyles. Socio-economic status is measured by three variables. They are woman's education, employment status, and religion. Education is expected to broaden individuals' perception of life by making them aware of various career opportunities and of the possibilities for self-fulfillment in many other ways. On the other hand, the time devoted to childbearing and childrearing may be perceived as competing against other opportunities. Women achieving a high educational level are expected to delay their entry into a first union, to be more knowledgeable and receptive about contraceptive methods, and to prefer smaller families, as compared to uneducated women. Religion is considered a socio-economic variable in this study and not a social background variable for the following reason. Women may change their religious affinity as they grow up. Reported religious denominations may therefore differ from those of parents, due to historical factors attached to the introduction of new religions in this setting. Basic demographic characteristics: Variables in this category are used as controls in the analyses. They are age at survey and number of children ever born (or parity). In a cross-sectional survey, women from different birth cohorts are interviewed at one point in time. These women have had different experiences in their lifetimes. Women belonging to the same birth cohort are likely to be influenced by the same events throughout their lifetime. Also, women at a certain parity may adopt behaviors and attitudes that are independent of their social background and socio-economic status. The effect of age and parity may disturb relationships to be studied. To sort out the effect of social background and socio-economic status on the dependent variables, it is important to hold the effect of age and parity constant. Due to the established correlation between age and parity, these two variables are not used simultaneously as controls in the models considered. Family formation and characteristics of unions: In this division, there are seven variables: Spouse choice, age at first union, type of union, type of marriage, number of unions ever contracted, spouse's residential status, and spouse's residence with extended family. Spouse choice indicates whether a woman chose her husband out of free choice or whether the choice was influenced by her parents. Type of union indicates whether the current union is a monogamous or polygynous. Type of marriage describes the form of marriage contracted -- legal, religious, or customary. Spouse's residential status indicates whether spouse lives in the same household or in a separate household. And residence with extended family indicates whether any of the family members of the couple live in the couple's household. This variable is constructed from responses to two questions. The first one is whether the wife has the responsibility to cook for any of the family members living the household. The second one is whether she has the same responsibility in regard to family members of the husband. The categories are: Some extended family and No extended family. Proximate determinants of fertility: Variables in this division are the known intermediary factors through which socio-economic variables affect fertility levels. In the analyses, emphasis is on breastfeeding and sexual abstinence. The two variables are divided into segments. The duration of breastfeeding is divided into two segments: the first one is reported durations up to 18 months and the other one is that of durations above 18 months. For the duration of sexual abstinence, the first segment is that of durations up to 12 months and the second one is that of durations above 12 months. These cut-off points are chosen taking into account norms of behavior and heaping problems. The idea in the analyses is to study differentials in characteristics associated with women practicing long durations and short durations. Women were also asked questions on the type of contraceptive method they are currently using, either traditional or modern. They were also asked whether they would consider induced abortion as an option if pregnant the month of the interview. Fertility Intentions: In this division are variables measuring women's intentions towards their own fertility and their impressions of their husbands' intention towards additional children. The variables are the following: desire for additional children, ideal family size, and husband's intention for additional children. These variables are indicators of the demand for fertility. 3.2 Statistical Methodology The statistical analyses follow two main steps: firstly bivariate analyses are conducted to study simple associations between independent and dependent variables and secondly multivariate analyses are conducted to sort out the effect of each independent variable on the dependent variable, holding constant other variables that might influence the relationship. It is necessary to investigate further bivariate relationships through a multivariate framework because simple bivariate associations are affected by the effect of other variables. In the analyses all dependent variables are dichotomies. They have the value of 1 in case of occurrence and that of 0 in case of non-occurrence. Usually, the ordinary least squares method is used in studying relationships between independent variables and a dependent variable in multiple regression analysis. In the case of a dichotomous dependent variable the logistic regression technique is more appropriate (Hosner and Lomeshow, 1989; Aldrich and Nelson, 1984). The model is estimated using the maximum likelihood procedures. The method maximizes the likelihood of the observed data. In other words, the method derives by several iterations the best parameters that fit the observed data. In studying the effect of social background and socio-economic status on alternative living arrangements, a special case of the logistic approach is considered. That is the case where the dependent variable is not a dichotomy but a polytomous variable. In this case, the basic principles of the logistic approach remain the same but the interpretation of coefficients is done with reference to a selected reference category of the dependent variable. 4. Summary of the Main Results This study indicates that the marriage institution, postpartum behavior, and fertility intentions are changing in urban Togo. Results suggest that while marriage remains universal, some customary marriage practices are being adapted to the urban environment. The participation of lineage members in the mate selection process, the payment of a bridewealth, and the practice of polygyny have resisted the legislation of colonial administrations, the evangelism efforts of missionaries, and Western education. Still, these customary practices have taken somewhat altered forms under the transformed social and economic order. As of the 1980s, the authority of lineage members in the mate selection process, although weakened, is expressed by close family members who manifest their approval or disapproval on the choice of a spouse. The customary marriage ceremony is also practiced, though less formally, along with religious and legal marriage ceremonies. Moreover, polygyny has survived the forces of change. Urban polygyny, however, is associated with various alternative living arrangements on the part of spouses. The tendency of spouses to live in separate residences has emerged as an important characteristic of unions in the city. As shown by Figure 1 which illustrates type of unions and living arrangements, spouses adopt a variety of living arrangements within unions. A woman in monogamous union can either live with her husbands or live separately. Similarly, a woman in a polygynous union can live in the same household with a co-wife, live alone with her husband, or live in a separate household. These living arrangements are further complicated with the tendency of spouses to live with some relatives or members of the extended family. From the study of social background and socio-economic status as related to living arrangements, results indicate that women from privileged background and high socio-economic status are more likely to live separately. This finding is consistent with historical and structural changes that have created a sexual division of labor that encourages female economic independence as well as a divergence in goals and priorities within the conjugal unit. At this point, with the nature of the data used in this study, it is not possible to state clearly the extent to which separate living of spouses is freely chosen or the result of constraints. Concerning the study of postpartum behavior in the 1980s, results indicate that the customary practices of prolonged breastfeeding and sexual abstinence, which traditionally sustained long durations between births, are now eroding in the urban areas compared to rural areas as well as among women with higher socio-economic status. The practice of sexual abstinence following birth seems the most affected. This is illustrated by Figure 2 which shows the proportion of births whose mothers are still abstaining falling earlier than that of births whose mothers are still breastfeeding. Results also suggest that the declining tendency observed in duration of breastfeeding and sexual abstinence is not being followed by substantial use of modern contraception. Traditional methods seem to be most preferred in this setting. This might be explained by prevalent myths wherein modern methods are viewed as devices that could permanently hinder women's childbearing abilities. These findings suggest a tendency towards increasing fertility, but other evidence seems to point in the opposite direction. As indicated by Figure 3 which shows the age-specific fertility rates from 1972 to 1988 in urban Togo, fertility might be decline in this setting taking results at their face value. Based on earlier results obtained in studying the age at first marriage, and the evidence of a low level of modern contraceptive use, the possibility exists that induced abortion might be playing an important role the observed decline of fertility. While ethnographic literature found evidence of this practice in some settings in Traditional Southern Togo, where induced abortion might be perceived as a contraceptive method. A this point however there is no solid quantitative evidence to support this claim. An alternative line of argument is that of traditional contraceptive methods depressing fertility in this setting. But similar to the argument suggesting induced abortion, there is no convincing evidence on the effectiveness of traditional contraceptive methods. Another area of investigation of this study is the demand of fertility. The examination of fertility intentions suggest that the majority of women in the urban area would like to space their children rather than to stop childbearing. Only women with secondary education were found to be a significant group likely to stop childbearing at relatively low parities. Results based on wives' reports indicate, overall, a relatively high level of agreement between spouses on the desire for additional children. Concerning ideal family size, women in both urban and rural areas reported preference for a family of four children. This tendency is generally shared by all subgroups of the urban area studied. This intention, however, is not being effectively translated into actual behavior. More than half of currently married women in the city were found in need of family planning, mostly to space their children but also in some cases to stop childbearing. 5. Relevance of Results to Theory From a theoretical standpoint, results of this study have demonstrated the advantages of approaching behavior from a pluralistic perspective wherein endogenous historical factors and contextual structural conditions and constraints are taken into account. Although only cross-sectional differences were examined, the patterns found suggest that convergence of behavior towards the nuclear family model as predicted by developmental theorists is not occurring in the majority of instances as of the 1980s. Instead, the evidence suggests the adaptation of customary patterns in the urban area. This is most apparent in the results on the institution of marriage. While the nuclear family model was expected, there is instead the emergence of alternative arrangements characterized by spouses living in separate residences. However, other changes in the marriage institution -- such as the weakening of lineage members' participation in the mate selection process -- are consistent with the predictions of developmental theorists. To a lesser extent, results indicating low levels of modern contraceptive use and the possibility of induced abortion following the erosion of the practices of breastfeeding and sexual abstinence, if proven correct, may also support the alternative of considering diversity in pathways of reproductive change. In sum, the results of this study support the non-developmental perspective wherein a variety of contextual institutional factors are considered in understanding behavior, rather than a preestablished pattern of expectations. It appears that theoretical work should be encouraged to conceptualize and provide measurement tools whereby individual behavior can be viewed as a function of contextual historical and structural conditions and constraints. Furthermore, conceptualization of how the socio-cultural environment itself changes and its patterns of transformation are important theoretical considerations. 6. Implications for Policy and Research The finding that some spouses live in separate residences suggests a growing tendency toward female-headed households, as compared to the past. This tendency is often perceived as a signal of the breakdown of the family institution. Female-headed households are generally found to be poorer than male-headed households. Based on this assumption, this finding is relevant for family welfare policies. Research is needed, however, to confirm whether female-headed households are effectively disadvantaged compared to male-headed households in this setting. Also, studies on the dynamic aspects of conjugal relationships and how they affect the emotional development of children are important research issues. Another finding relevant to policy and research is the declining tendency in the practice of breastfeeding. Besides the effect of nursing in regulating the duration of the postpartum infecundable period, breast milk has immunological properties that protect the infant from many illnesses such as diarrhea and respiratory-tract infections. This protection is lost in the case of improper supplementation and may affect infant and child health. Programs explaining the advantages of breast milk to the health of infants appear relevant, as does research to document the impact of supplemental foods on infant and child health. In addition, the myths surrounding modern contraception should be further explored in order to design programs to clear them out. It appears that family planning programs in this setting should concentrate further on explaining how modern contraceptive devices affect the physiology of the user. They should also consider the difference between needs for spacing and for stopping in the supply of methods. Most importantly, the modern health system of family-planning service delivery should be more sensitive to the social and religious background of clients. Finally, a thorough study of fertility levels and trends and their determinants is needed. This investigation should evaluate the impact of induced abortion and spacing behavior on fertility. Studies are also needed to assess whether these are convincing signs of a beginning of fertility decline in the country and to evaluate the respective influences of adaptive and innovative behaviors. NOTES
REFERENCES
Copyright 1994 - Union for African Population Studies. |
| |||||||||
