+Bioline International Official Site (site up-dated regularly)

African Population Studies/Etude de la Population Africaine, Vol. 20, No. 1, 2005, pp. 19-37

Given the recent increase in the prevalence of cohabiting unions in Botswana, this paper speculates on the role this type of union plays in contemporary Botswana family system. The analysis is based on the combination of qualitative and quantitative data from the 2001 Cohabitation Survey and the 1996 Botswana Family Health Survey respectively. The overall results reveal that cohabitation differs from marriage in terms of socio-legal status and the reproductive behaviour of women in cohabiting unions. The paper further presents evidence that, of the three common interpretations of cohabitation - an alternative to marriage, a temporary phase before marriage and an alternative to being single - cohabitation in Botswana can be viewed as a temporary phase before marriage. Against this evidence, the paper concludes that while the increasing prevalence of cohabitation does not threaten the institution of marriage in Botswana, it results in delays in timing of marriage.

Nuptiality patterns in Botswana have witnessed considerable changes over the past three decades, with trends showing significant increases in proportion never married and average age at marriage (Gaisie, 1995; Mukamaambo, 1995; Mookodi, 2003). Another notable recent trend is the increase in the proportion of people who live with partners of the opposite sex in sexually intimate relationships without being legally married—a living arrangement known, inter alia, as cohabitation, cohabiting unions, consensual unions or living together. For example, according to the 1991[1] census of Botswana, 12% of all people aged 15 years and above reported themselves as cohabiting. By the time of the 2001 census, this figure had increased to approximately 17%. The census results also show that the proportion of current cohabiting unions relative to all current unions increased from 31 to 46% between 1991 and 2001. In the same vein, data from the last two Botswana Family Health Surveys (BFHS) revealed that the proportion of women aged 15-49 years who were in cohabiting unions increased from 11 to 17% between 1988 and 1996, while proportion of cohabiting unions among all unions increased from 28 to 50% between the two time periods (Mokomane, 2004).

The increasing prevalence of cohabitation in Botswana has catalysed concern among policy makers (for example, Women’s Affairs Department, 1999) and social scientists (for example, Molokomme, 1990; Meekers, 1991; van de Walle, 1993; Lesetedi and Ncgoncgo, 1995;) and has contributed to their calling for comprehensive research to examine this phenomenon within the broader societal context. As a partial answer to their call, this paper examines the meaning or role of cohabitation in Botswana. Although this is a complex and rather ambiguous question (Smock, 2000:7), cohabitation researchers have posed three possibilities of viewing this type of union, all of which can affect the interpretation of the changes in marriage (Manting, 1994).

One view is that cohabitation is an alternative form of marriage (Manting, 1994; Carmichael, 1995; Smock, 2000), and it may reflect a rejection of marriage as an institution, or it can be a true alternative (Prinz, 1995:76). Viewed in this way, an increase in cohabitation is a threat to the institution of marriage and may play a major role in the decline in the prevalence of marriage (Smock, 2000). This view can be partly explained by the process of individualism, where there is greater need for flexibility, individual freedom and independence, especially among women (Gage-Brandon, 1993; Meekers, 1993; Manting, 1994). For example, writing about the gradual increase of consensual unions observed in many African societies, Meekers (1993:35) stated that:

In traditional bridewealth marriages, husbands have authority; husbands expect their wives to be obedient, and they tend to make claims on their wives’ labor and income… Hence, women’s desire to gain status through economic independence is often a source of conflict within the union. In an attempt to avoid such conflicts a growing group of women now try to escape male control by steering clear from bridewealth, marriages…Rather than contracting a formal marriage, these women prefer unmarried cohabitation or prefer to have lovers who do not live with them because this allows them to maintain liberty.

The second view is that cohabitation is the last and temporary phase before marriage (Manting, 1994; Carmichael, 1995; Smock, 2000). This view implies that cohabitation is a transitional stage that is either terminated or transformed into a legal marriage (Wiersma, 1983 cited in Prinz, 1995: 77). That is, cohabitation is seen as a trial period and a major reason for the delay, but not an overall decline, of marriage. Thus its emergence does not threaten the institution of marriage (Manting, 1994). This is mainly because, by living together unmarried, cohabiting couples have the opportunity to know each other in daily life situations or test their compatibility. Those unions that are deemed successful are transformed into marriage whereas others are dissolved (Prinz, 1995: 77). For example, in many sub-Saharan African societies where men usually postpone a formal marriage until they have proof of their prospective wives’ fertility, cohabitation can be “considered a trial marriage during which pregnancy becomes a means of testing the relationship” (Meekers, 1991:2). If the trial is experienced as successful, they marry; if not, they break up (Meekers, 1991:2).

Finally, cohabitation can be seen as an alternative to being single (Rindfuss and VandenHeuvel, 1990). Viewed in this way, it “represents an extension of the dating and sexual relationships and its ideology does not include permanence” (Smock, 2000:8). That is,

At the onset of the process, marriage is not even considered; the decision to cohabit is only a consequence of the time one shares with each other. Independence is highly valued, whereas commitment has a rather low value (Manting, 1992 cited in Prinz, 1995:78).

According to this view, a rise in the prevalence of cohabitation implies a greater number of sexual partners for individuals as well as an increase in union dissolution rates (Rindfuss and VandenHeuvel, 1990). Quite possibly, then, this could have a feedback effect on the institution of marriage by weakening its central foundation of permanence (Rindfuss and VandenHeuvel, 1990:722).

The legal position of cohabiting couples in Botswana is, to a large extent, dictated by the prevailing policy in the country that marriage should be encouraged (Otlhogile, 1994:3). Consequently, not only is cohabitation unrecognised as an institution by the two systems of law that operate simultaneously in the country (general and customary law), but neither of the two systems gives cohabitants any legal protection. This is particularly so with regard to inheritance and property rights of cohabiting women as well as the maintenance of children born within these unions. Under customary law, for example, “no length of cohabitation is considered to amount to marriage or give rise to inheritance rights between the partners or their issue” (Dow and Kidd, 1994:32).

By the same token, general law does not impose a duty of support between a cohabiting couple, nor does it imply joint ownership of property they acquire over the duration of the union (Molokomme, 1990; Alexander et al., 1992). Therefore, if a cohabiting man dies without leaving a will, his partner cannot inherit intestate succession (Otlhogile, 1994). According to Molokomme (1990:19), apart from the Workmen’s Compensation Act (Cap 47:03), which gives a woman who can prove that she was dependant on a deceased workman a right to claim compensation, cohabiting women have no legal protection:

[Cohabiting] women often end up in a precarious position in relation to property that the couple acquires during the subsistence of their relationship. Usually, the woman in such a relationship is either a housewife or earns a small wage outside the home, while the man will earn more because of the relatively better employment opportunities for men in the formal sector. … As a result, whatever property of value is acquired during their friendship will be in his name. Since the law does not recognize such a relationship as giving rise to duties of support between them, or as implying joint ownership of property they acquire, women are left without support or property should the friendship terminate. … as a general rule, women who cohabit with men without being married to them have no legal protection (Molokomme, 1990:13).

In the same vein, children born to cohabiting couples are perceived, both in fact and in law, as non-marital. Consequently if their parents break-up, they are subject to the same maintenance laws as children born to unmarried mothers. A discussion of these laws is beyond the scope of this paper (see instead, Molokomme, 1990; Alexander et al., 1992; Kebonang and Lebotse, 1999). However, these laws are succinctly described in the following statement by Dow and Kidd (1994:104):

In a culture which assumes that whenever a man and a woman have established a common household, all the property accumulated belongs to the man, the successful challenge of the validity of a marriage will almost always mean loss of property of the union by the women and her children. No length of cohabitation will make her entitled to those resources. Marriage is the only recognized way a woman may access resources accumulated by the father of her children, for herself and the children.

Societal attitudes towards cohabitation are equally negative, with qualitative evidence (Mokomane, 2004) showing that marriage is perceived by the society as the more serious and ideal type of union, while cohabitation is treated trivially.

This paper uses a combination of qualitative and quantitative data. The qualitative data come from a survey (hereafter called the 2001 Cohabitation Survey) conducted in Botswana between September 2001 and March 2002. The survey addressed an array of issues concerning cohabitation as an increasing phenomenon in Botswana. The data were, however, not meant to be statistically representative but, rather, to enable an in-depth analysis of the issue under study. Thus, to compare rural and urban perceptions, the survey was conducted in only two localities—the urban town of Lobatse and a rural village in the Ngwaketse district. Lobatse’s 2001 de facto population of 29, 689 and socio-economic facilities and activities make the town cosmopolitan enough to give insights into the views and behaviour of urban dwellers in Botswana. Similarly, the chosen rural village can be said to be typical of other villages in Botswana in terms of social organisation, political system, system of administration and justice as well as availability of social services.

The 2001 Cohabitation Survey used a variety of data collection methods such as in-depth interviews, focus group discussions, key informant interviews and court records examination. However, for the purposes of this paper, only excerpts from the 36[2] in-depth interviews conducted with male and female cohabitants in the survey are used. The 2001 cohabitation survey defined a cohabitant was a person who had been living together outside marriage[3] for at least six months. The six months or more duration criterion was placed on cohabiting unions primarily to leave out episodic and short-term relationships.

The analysis of the qualitative data adopts an issue-focused approach. This approach, described by Weiss (1994:154) as an analysis that concerns “itself with what could be learned about specific issues—or events or processes—from any and all respondents”, involves four distinct analytical processes: coding, sorting, local integration and inclusive integration. The coding generally entails identifying themes contained in specific text passages or segments (Gorden, 1992; Miles and Huberman, 1994). Sorting, on the other hand, involves the categorising of the (coded) data into separate and appropriate excerpt files while local integration summarises the excerpt files and their coding hence bringing coherence and meaning to the data. Inclusive integration finally “knits into a single coherent story the otherwise isolated areas of analysis that result from local integration” (Weiss, 1994:160).

The Ethnograph computer software was utilised in undertaking the above process. After transferring all the transcripts from a word processor, the coding entailed giving text passages containing identical themes similar codes, while those containing different themes were coded differently. The major themes were the respondents’ background information, their reasons for the establishment of the union, attitudes towards cohabitation and marriage and, for women, reproductive behaviour. After the coding, the data were sorted by extracting excerpts dealing with the same issue, which were used to create excerpt files. It is these files that were constantly used and referred to during the writing up of the results

The present analysis relies heavily on verbatim reporting and throughout the paper the excerpts are presented using the “standardised approach” (Weiss, 1994:193). That is, while material presented remains “true to the words and meanings of the original”, minor editing was occasionally done to make it easier for the reader to grasp. For example, informal spacers and false starts such as “uh” and “um” were dropped from all excerpts used. In no instance, however, were words that were never said by a respondent used. In the few instances where a word or phrase was deemed necessary to clarify a meaning, the word or phrase appears in parentheses.

The quantitative data utilized come from the 1996 Botswana Family Health Survey (BFHS III). The BFHS III was taken under the Demographic and Health Surveys (DHS) program and it had the broad objectives of providing up-to-date information on family planning awareness, approval and use as well as basic indicators of maternal and child health and other topics related to family health (Central Statistics Office, 2004). In addition, the survey collected information needed to explore trends in fertility and mortality and to examine factors influencing these basic demographic indicators. The focus was, therefore, on women of reproductive age; no information was collected from men. The BFHS III identified 8,895 eligible women aged 15-49 years of which 8,483 or 95% were successfully interviewed (Central Statistics Office, 1999). For the present purposes, the BFHS III data are used to quantitatively examine, using bivariate and multivariate techniques, the fertility behaviour and preferences of currently cohabiting women relative to their married counterparts. The multivariate technique used is the binomial logistic regression—a form of regression used when the dependent variable is dichotomous (has only two values) and the independent variables are of any type. For a more detailed description of this technique see, among others, Polit (1996) and http://www2.chass.ncsu.edu/garson/pa765/logistic.htm .

To set the stage for the presentation of the results, Table 1 shows selected basic socio-economic characteristics of cohabitants in Botswana relative to their married counterparts.

The results suggest that cohabitants in Botswana are, in general, socio-economically disadvantaged. That is, not only are they, on average, younger than their married counterparts, but they are also less educated, less religiously involved, more likely to be unemployed, more likely to work in elementary (blue-collar) occupations and more likely to live in urban areas than married people. While this finding departs from the general pattern noted in other sub-Saharan Africa countries (Calves and Meekers, 1999; Gage-Brandon, 1993), where these unions are more prevalent among the economically independent and better educated, it is consistent with the pattern in many Latin American countries (Castro Martin, 2002). Taken as a proxy for socio-economic status, lower education and unemployment implies limited resources to establish legal marriages, thus making cohabitation the next best alternative.

Indeed, Mokomane (2004) found that cultural and structural changes that create constraints to marriage underlie the increased prevalence of cohabitation in Botswana. Along with the decline in traditional morality and the diminished gerontocratic control, a salient cultural change relates to transformations in the customary marriage procedure, specifically the monetary value that has now been placed on formerly symbolic gestures of a customary Tswana marriage. In particular, the inflated and stringent demands placed on contemporary bridewealth means that many unemployed or lowly paid young men have difficulty in meeting the costs associated with getting married.

Regarding structural changes, unemployment transpired as one of the main factors that have facilitated the increased levels of cohabitation. Mokomane’s results showed that getting married in contemporary Botswana society is more feasible for those people who have relatively higher incomes, are employed or at least have access to credit facilities. Thus, given that most cohabitants are more likely to be unemployed and young (hence more likely to be at entry-level incomes and less likely to have investments), the total costs of getting married are prohibitive and can be responsible for the increase in cohabitation.

One way of speculating about the role of cohabitation in a population is to establish whether cohabitants have rejected marriage or whether they still intend to marry in the future (Khoo, 1987:189). Another way is to compare how childbearing patterns vary among cohabiting and married women (Smock, 2000:8). According to Calves and Meekers (1999:618), the difference in the fertility patterns can clarify whether cohabitation is conceptually different from marriage, or whether it should be considered merely a variation of the normative marriage. For example, if cohabiting unions are an alternative to marriage, it is expected that there will be no significant differences in fertility according to type of union (Meekers, 1991: 3). On the other hand, if cohabiting unions are a transitory stage in the married life cycle, the instability associated with the unions will result in lower levels of exposure, and consequently, in lower levels of fertility than in the case for legal marriages. This subsection will therefore begin by presenting cohabitants’ attitudes towards marriage and cohabitation. This will be followed by a comparison of their fertility behavior and preferences relative to those of married women.

All cohabiting respondents in the 2001 Cohabitation Survey were asked whether they wanted to marry or had any intention of getting married in the future. Consistent with previous findings by other researchers in Botswana (for example, Molokomme, 1991; Pitso, 1997; Mokomane et al., forthcoming), all respondents, except one woman who had a problem with the age difference between her and her partner (she was ten years older), articulated a strong desire for marriage. Indeed, when asked about the perceived advantages and disadvantages of cohabitation compared to marriage, almost none mentioned any advantage of cohabitation. The most frequently cited disadvantages of this type of union, as compared to marriage, were lack of security in such unions as well as the lack of legal protection if the union was terminated. The following quotations typify this apprehension:

It [cohabitation] is not good in that in the event of the death of one of us, you might find that the deceased’s relatives are very selfish. For example, if I die and my relatives are selfish, they can come and take everything in this house and by so doing denying the children I have with her any inheritance. All on the basis of the fact that I had not married her (45 year old cohabiting man, rural area),

Cohabitation has one major disadvantage: you see I am currently living with this woman and we buy things jointly and pay the rent and utilities together. If I can die the major problem is going to be that my parents will come rushing and if they like they will just say to her parents “ you daughter never bought anything, all these things are ours”… (31 year-old cohabiting male, urban area)

Cohabitants thus felt that marriage, unlike cohabitation, gives a sense of legal protection, stability as well as some degree of security for both the cohabitants and their children. For example:

Marriage is important for security; you know without doubt that that this person [the partner] is mine. My children will also grow up knowing that they are being raised in a family of married people (35 year old, cohabiting man, urban area).

I want to get married because then if he abuses or mistreats me I would have rights and legal protection. Currently I am just living with him, bearing children for him but he can go and come as he wants and there is little I can do (25 year old cohabiting woman, rural area).

I want to secure my children’s future so that in the event of his death everybody should know that his property belongs to (our) children and no one can come and claim them. Now if you are not married, he cannot even put us down as beneficiaries [for his terminal benefits] at his place of work (27 year old, cohabiting woman, rural area).

For some cohabitants the aspiration for marriage is based on the conviction that marriage brings respectability and also guarantees some social acceptance and support from in-laws especially if the couple has problems:

I would like to get married because marriage gives one some degree of dignity, you don’t lose your dignity by having multiple relationships that end within months (36 year old cohabiting woman, urban area)

Marriage is important because you build a solid and respectable home. Wherever you are, you have dignity and respect from other people (32 year-old cohabiting man, urban area)

I want to have a respectable home where I can live with a husband and children. I don’t want to be called a hawk where other women will always think that I will snatch their husbands. I don’t want to be called names (30 year old cohabiting woman, urban area)

The good thing about marriage is that your wife’s family knows you and you are welcome to go there anytime. If you have problems with her you can talk to her parents. Similarly if she has problems with you she can talk to your parents. If you have problems it becomes your problem and you face that together with your parents. Now if you are not married, that is a problem… (32 year old cohabiting man, urban area)

By the same token, a 38 year-old cohabiting man in the rural area explained his reasons for desiring marriage as follows:

Firstly, the reason I want to get married is that in this village of ours a lot of people are getting married. My relatives and my nieces and nephews are also getting married. Now instead of me, as a maternal uncle[4], taking part in the certain rituals I have to ask some married men to do that on my behalf. Now this thing worries me. Because I am not married I cannot even perform my traditional chores. When my nephews or nieces get married I have to be in the lead; now I have to ask any man to act as their maternal uncle and he will take all the gifts that are due to me. Besides that, I find it really embarrassing. In the past I did not realise these things because I used to live and work in South Africa. Now ever since I came back home and observed the traditional way of doing things, I have realised that I am almost a ‘nobody’.

Although the majority of the cohabitants expressed a high level of uncertainty about the timing of their marriages, with many stating that they do not see themselves doing so until well in the future, the foregoing findings suggest that, as in the traditional society, marriage as an institution is still placed in high esteem in contemporary Botswana. The dominant theme that reverberates throughout all the transcripts regarding the attitude towards cohabitation and marriage is that marriage, unlike cohabitation, offers greater legal securities, stability and social acceptance and support than cohabitation.

Table 2 shows the distribution of cohabiting and married women aged 15-49 years by selected indicators of reproductive behaviour.

According to the table, although an overwhelming majority (over 90% of both groups of women) had at least one birth at the time of the BFHS III, cohabitants were more likely than married women to be childless (9.7 and 3.1% respectively). The table also shows that cohabitants had a considerably lower mean number of children (cumulative fertility) compared to their married counterparts. It is noteworthy that although its historical nature may render it irrelevant to the current situation, cumulative fertility is useful in examining how average family size varies across groups. This, in turn, is important for understanding current fertility (Central Statistics Office, 1999:10). For example, for the present purposes it shows that the fertility patterns in the two types of unions under study are not entirely similar.

With regard to fertility preferences, Table 2 shows that the average ideal family size was also lower for cohabiting women compared to those who were married. The desire to have a child in the next two years has been held to be a commonly used indicator of fertility preference that is closely related to current union type (Castro Martin, 2002:45). That there was virtually no difference between married and cohabiting in terms of wanting a child within that time period is inconsistent with a common theory that cohabiting women may want to demonstrate their fertility in order to speed up proceedings for a legal marriage or use childbearing as a means of getting a man’s economic resources (Meekers, 1991).

Indeed, multivariate analysis shows that after controlling for selected socio-economic and demographic characteristics, cohabitants were, in fact, significantly less likely to want a child in the next two years than married women (Table 3). This can be partly explained by the fact that because a considerable proportion of cohabiting women are mothers (Table 2), their fertility or ability to give birth is already proven. Strong evidence from the qualitative survey also showed that for most cohabiting women the desire for additional children was conditional upon firmer commitment from partners in terms of marriage. Statements made by these women revealed that the majority of them were implicitly aware that their unions are not only unstable but the legal constraints of child maintenance for unmarried mothers are likely to leave them, and any children resulting from the unions, with little legal or financial support from the fathers if the union was terminated. For example:

I can only have child if I get married and my husband wants a child. Otherwise I will never have another child out of wedlock because that is just putting yourself in more struggles (31 year old cohabiting mother of four; junior secondary school education; housemaid; urban area).

I can only have one more child in future but that will only be after getting married. Unless I get married I will not want another child for the rest of my life. If I don’t get married, I will never have another child at all. This one (that I have) would be enough because a child is difficult to raise (31 year old cohabiting mother of one; senior secondary school education; public servant; urban area).

He often talks about us having a child but I have told him that I will not have a child until he marries me. I tell him that ‘I know that you want me to have a child and afterwards you will dump me. So I am not having a child until you marry me’ (25 year old cohabiting mother of one; unemployed, rural area).

Given that descriptive analysis of the BFHS III shows that only 2.9% of cohabiting mothers were receiving support from their children’s fathers through the Affiliations Proceedings Act, it is therefore probable that these women have experienced abandonment and the responsibility of looking after children with little or no support from the fathers before. The general perception, among the women was, therefore, that child-rearing is easier and more secure when the couple is married because then the fathers are more obliged to maintain their children.

The overall results in the foregoing subsection suggest that cohabiting unions in Botswana do not correspond to the notion of cohabitation as a true alternative to marriage. Firstly, the negative socio-legal attitudes to cohabitation in the country discussed earlier indicate that the two unions cannot be considered similar or equal. Secondly, contrary to the argument that if viewed as an alternative to marriage, cohabitation can be a rejection of marriage, the above results show that for virtually all the cohabiting respondents in the qualitative survey, marriage had not been rejected at all. Indeed it was the most desirable form of union because of its higher stability and the better defined socio-economic and legal rights of spouses and their children. In particular, cohabiting women did not consider their unions as equivalent to marriage. They recognized that unlike their married counterparts, they had a higher risk of being left to raise children alone with little legal or financial support from the father if the union breaks up. Furthermore, in addition to the clearly different reproductive behaviour and preferences of cohabiting and married women, cohabitants’ arguments that marriage offers the best childbearing environment indicate that the two unions are indeed different.

It seems, from the results that cohabitation in Botswana cannot be viewed as an alternative to being single either. As stated earlier, when cohabitation is seen as an alternative to being single, the necessity of a long-term commitment does not exist and the union “does not assume a commitment to permanency at the beginning of the relationship” (Rindfuss and VandenHeuvel, 1990:707). However, not only did the results in this study show that the majority of cohabitants in Botswana want to get married, but Mokomane (2004:130) also showed that the majority of the cohabitants are ‘purposeful delayers’. That is, they established their cohabiting unions after dating for more than one year and after discussing and reaching some degree of informal understanding about the conditions of the union. One of these conditions was usually that marriage would be the next logical next step after the cohabiting union (Mokomane, 2004:130)

Although the degree to which cohabiting unions are transformed into marriage cannot be determined with the data at hand, of all the three possible interpretations of cohabitation, the view that cohabitation is a temporary phase before marriage bests fits the empirical data presented. The results clearly show that in Botswana marriage is the “ultimate life goal” for virtually all cohabitants while cohabitation is a compromise living arrangement for those who are not yet ready for marriage, and may not be for a considerable period of time, mainly due to economic reasons (Mokomane, 2004). For many cohabitants, marriage is a distant goal that will be attained once the current constraints are overcome. In addition, Mokomane (2004) revealed an inverse relationship between age and the proportions of cohabitants, a finding she attributed partly to a gradual termination of these unions through legalization as they mature.

If cohabitation is a prelude to marriage, as the findings in this paper suggest, then there is little evidence that increasing cohabitation threatens the institution of marriage in Botswana. This is particularly so given that the fact that there was no evidence of a rejection of marriage by cohabitants, and that the vast majority of them expressed a great desire to marry and/or expect to eventually marry. Marriage remains the most desirable union because of its higher security benefits, its symbolic commitment of endurance and a more clearly defined social position with respect to the partner and the in-law family. Cohabitation, on the other hand, can therefore be viewed as a temporary convenient arrangement that will be left once the constraints to marriage are removed.

While not every one who wishes to marry will do so (Tanfer, 1987:494), there is no indication that cohabitants are more likely to choose this lifestyle as a permanent way of life. Certainly, economic insecurity (see Mokomane, 2004), at least in the short run, might accentuate the increasing prevalence of cohabitation. However, the fact that most people eventually seek socio-economic and legal security in relationships, which cohabitation does not provide, there is little chance that cohabitation will pose a widespread challenge to marriage as an institution, at least in the foreseeable future. To the extent that cohabitants will reach their marriage ideal, it can be concluded from the evidence in this paper that increasing cohabitation levels will not result in any sharp decline in the proportion of people who will ever marry because the overwhelming majority of cohabitants do plan to eventually marry. However, because cohabitation tends to delay marriage, its influence in the marriage patterns should be seen in (increasing) higher ages at first marriage.

What are the policy and programmes implication of the findings of this paper? The results suggest that cohabitation in Botswana may not be a simple ephemeral phenomenon. More and more young people are likely to experience cohabitation before eventually marrying while, given the high prevalence of non-marital childbearing, the proportion of children who will ever live with cohabiting parents is also likely to increase over the years. There is, therefore, an urgent need to re-examine our social and legal attitudes to cohabiting unions. In particular, there is need to explore the best ways to protect, in the event of union dissolution the property, inheritance and maintenance rights of cohabiting women and their children, without undermining the institution of marriage. As Otlhogile (1994:13) argued:

“…relationships outside marriage are a reality and we cannot afford to ignore them any more”. … The imposition of rights and responsibilities between cohabitees will help curb other social malaise. It will reduce or prevent “illegitimate births”.

Against emerging literature showing the deleterious effects of cohabitation on children (Brown, 2002; Manning, 2002), there is also need to qualitatively examine, in the context of Botswana, the impact of this union type on children’s psychological, emotional, behavioural and cognitive outcomes.

[1] The Central Statistics Office only started to enumerate cohabiting unions as a separate marital status category (living together) in 1991. Pitso (1997:111) and Mokomane (2004:85) argue that before then, cohabitants are likely to have been classified as ‘never-married’ or ‘married’.

[2] Of these, 21 were with cohabiting women (15 in the urban area, 6 in the rural area) and 18 (11 in the urban area, 7 in the rural area) were with married women.

[3] For the purposes of the 2001 Cohabitation Survey, a marriage was defined as any union that had been solemnized by a statutory marriage officer and/or in which the traditional ritual/ceremony known as patlo - where the families of a prospective couple agree to the marriage - had taken place.

[4] In traditional Tswana marriage, the maternal uncle plays the most important part in his nephews and nieces’ weddings. However, like everybody else who takes part in the important marriage negotiations, he must be married. Otherwise, a married person of his choice must take his place (see Schapera, 1938).

The following images related to this document are available:

Photo images