Abstract
Migrants form nonfamilial ties in urban destinations, which could compete with origin families for a share of remittances. A framework of competing commitment predicts that new relationships affect remittances depending on the extent to which they substitute for the benefits provided by origin families. Analyses of data from urban migrants in Kenya show that serious nonmarital sexual partners substitute for psychosocial support from the rural family and that material transfers migrants give to these partners significantly reduce remittances. The findings have implications for the ways scholars conceive of competition, the nature of exchange, and substitution of support across intimate relationships.
Migration is often conceptualized as a process of social exchange. Particularly in developing countries, migration is frequently undertaken as a household strategy that generates streams of benefits and costs for both migrants and their families in the origin. Research from across the globe has documented how migrants receive multiple forms of social and economic support from ties in the origin, such as employment information, emotional encouragement, and the promise of family inheritance. In this context of reciprocal exchange, remittances are a major return obligation for migrants. Indeed, large proportions of migrants’ incomes are devoted to remittances, and these resources are often critical to reducing economic vulnerability of the origin family (Rempel and Lobdell 1978; Itzigsohn 1995; de la Brière et al. 2002; Azam and Gubert 2006).
Given the emphasis on origin networks as sources of support for migrants, the accompanying literature naturally focuses on these continuing ties as determinants of remittance flows. This research commonly concludes that stronger social ties to families and home communities increase migrants’ propensity to remit as well as the level of pecuniary assistance they provide (Hoddinott 1994; Menjivar et al. 1998; Mooney 2003; VanWey 2004; Piotrowski 2006). Remarkably absent from the literature on remittance behavior is the study of new nonfamilial ties that migrants forge in the destination and the benefits and costs associated with these exchange relationships. New ties provide various forms of assistance to urban migrants that could substitute for support from rural relations, while at the same time, they require reciprocal commitments, which are often financial or material in nature. In this case, new ties in the city compete for migrants’ scarce resources and could have significant repercussions for resource commitments to families in the origin.
The competing notions of familial and nonfamilial ties are particularly relevant to the internal migration experience in sub-Saharan Africa. On the one hand, African migrants have been described as living in a “dual system”: embedded in the new urban environment while continuing to identify and maintain close relationships with their rural households and kinship groups (Gugler 1991; Cliggett 2003). These familial ties have taken on new importance in the era of HIV/AIDS, as rural households provide the bulk of care and support for ill individuals, a large proportion of whom are returning migrants infected in urban areas (Clark et al. 2007). On the other hand, the HIV/AIDS epidemic has focused attention on a particular form of nonfamilial tie prevalent in urban Africa—nonmarital sexual relationships—and the pivotal role these partnerships have played in the spread of the epidemic (Caldwell, Anarfi, and Caldwell 1997; Brockerhoff and Biddlecom 1999; Lurie et al. 2003). Nonmarital sexual relationships are common and encompass a range of partnership types, including commercial sex, casual sex, and longer-term relationships with serious girlfriends. Past work has shown that these diverse partnerships also embody reciprocal exchange; they require substantial financial commitments on the part of male partners and provide various forms of social and sexual support in return (Castle and Konate 2003; Luke 2003; Kaufman and Stavrou 2004; Maganja et al. 2007). In light of these realities, African migrants must balance resource obligations between new nonfamilial ties in the city and existing familial ones in the rural origin. Do sexual partners and the money and gifts given to them affect remittances to the rural family? Considering the heterogeneity of sexual partnerships, where would we expect to see trade-offs and why?
I develop a theoretical framework of competing commitments to understand how individuals negotiate conflicting claims on their resources and predict where, why, and to what extent they make trade-offs across relationships. I build on the notion of substitution as conceived by several theoretical perspectives, which argues that goods, services, or intimate relationships can substitute for one another to the extent that they duplicate the same benefits. I recognize, in addition, that the support individuals receive in social relationships is valuable to them, and the more benefits they obtain, the greater resources they commit in return. It follows that when a new relationship is formed and strengthened, the increase in resources devoted to it will be associated with a decline in the strength of an existing relationship and, hence, accompanying resource commitments. The strength of this negative association between resources allocated to new and existing ties depends on the degree of substitution. At one extreme, engagement with a perfect substitute will lead to termination of the existing relationship and all prior resource obligations. In the case of a partial substitute, which duplicates only one of several types of support, transfers to the new relationship will partially decrease resource commitments and weaken the existing tie. For the case in which there is no substitutability across relationships, no trade-off in resources should be observed.
In sub-Saharan Africa, a variety of new and existing ties in the destination could duplicate the economic and social benefits typically supplied by the rural origin. Given the prevalence of nonmarital sexual partners in urban settings and the significant amount of money and gifts (what I refer to as “transfers”) devoted to them, an intriguing question is the role these relationships could play as alternative sources of support in the city. Nonmarital relationships provide a variety of forms of support to male partners, and therefore, some may be closer substitutes for the family than others. Throughout Africa, commercial sex and casual partnerships are generally brief encounters intended for sexual gratification. Longer-term serious girlfriends, in contrast, provide more than just sex: they offer emotional encouragement, companionship, and affection in the city (Cornwall 2002; Hunter 2002), which could replicate the family’s psychosocial support. The implication of my framework of competing commitments is that transfers to these serious sexual partners could reduce remittances to rural families. Furthermore, transfers to other types of sexual partners should have no association with the level of family remittances.
To test my predictions, I use quantitative and qualitative data collected from male migrants in Kisumu, the capital of Nyanza Province in Western Kenya and the third-largest city in the country. Kisumu represents one of a multitude of smaller African cities that serve as important interim or final destinations for large streams of internal migrants. The survey gathered information on the value of migrants’ remittances to their rural families and home communities as well as measures of ties to these relations. Kisumu is also the epicenter of a mature HIV/AIDS epidemic, with HIV prevalence reaching 25% or more of the adult population over the last decade (Glynn et al. 2001; NASCOP 2005). The survey and in-depth interviews were consequently designed with particular attention to men’s nonmarital sexual behavior, and both are among the first to elicit details of diverse nonmarital partners and the material transfers given to them. I use these unique data on migrants’ multiple expenditures to test my hypotheses regarding resource trade-offs between sexual relationships in the city and remittances to the rural origin. Furthermore, I explore alternative explanations for the negative transfers-remittances association, including the possibility that migrants’ spouses are the target of substitution for serious nonmarital partners, not the rural family, as well as the possibility that unobserved individual characteristics jointly determine transfers and remittances, giving rise to a spurious negative correlation. The series of results that I present support the theoretical framework and suggest that there is indeed substitution across competing commitments in Kenya—trade-offs that have important implications for resource distribution and support in this area of high HIV/AIDS prevalence.
BACKGROUND LITERATURE
Remittances Research in Africa
The theoretical and empirical research on remittance behavior is rooted in an exchange perspective and recognizes that migrants maintain relationships with their families of origin after they arrive in the destination.2 Both parties act to sustain these relationships not only out of purely altruistic motivations but because each can obtain valued resources from these established connections (Lucas and Stark 1985; Stark and Lucas 1988). In the African context and among the patrilineal Luo of Nyanza Province, the rural extended family and its residential compound include parents, spouses, and children as well as male siblings and their families (Ocholla-Ayayo 1976; Nyambedha, Wandibba, and Aagaard-Hansen 2003).
The reciprocal relationships between migrants and their rural families are referred to as “contractual” arrangements in the remittances literature and are divided into two main forms: coinsurance and investment strategies. The coinsurance strategy is aimed at diversifying risk for migrants and their rural households in the shorter term. Because insurance and credit markets are incomplete in many developing country settings, migrants and rural households rely on each other for material support in times of unforeseen “shocks” or hardships, such as periodic drought or crop failures in the origin and bouts of unemployment or temporary illness for migrants in the destination (Lucas and Stark 1985). In contrast, the investment strategy represents attempts to smooth longer-term consumption patterns through intertemporal, intergenerational arrangements. Here, the direction of resource flow depends on migrants’ stage in the life cycle. Migrants send remittances during their productive years in the city to repay families for investing in their education and to secure a portion of their inheritance, which will support the migrants in old age (Hoddinott 1994).
Several formative studies of remittance behavior were undertaken in the 1960s and 1970s in sub-Saharan Africa (Lucas and Stark 1985; Stark and Lucas 1988) and in Kenya in particular (Johnson and Whitelaw 1974; Rempel and Lobdell 1978; Knowles and Anker 1981). This research revealed the vast amounts of remittances that flow to rural families—between 6% and 30% of urban migrants’ earnings (Hoddinott 1994; see also Findley 1997)—and evidence of both types of contractual arrangements with kin. Overall, this work finds that migrants who maintain stronger ties with their origin families send larger remittances, similar to other remittances work globally (e.g., Menjivar et al. 1998; de la Brière et al. 2002; VanWey 2004; Sana 2005; Azam and Gubert 2006; Piotrowski 2006).
In the contemporary era, the HIV/AIDS epidemic has dramatically altered traditional patterns of income distribution and social support across Africa (Hosegood et al. 2007; Parker and Short 2009). High urban HIV prevalence has produced a reverse flow of infected migrants who, in the absence of formal health insurance and services, depend on their rural families for physical, emotional, and financial support during their illnesses and for the spouses and children they leave behind after death (Nyambedha et al. 2003; Chimwaza and Watkins 2004; Clark et al. 2007; Schatz and Ogunmefun 2007).3 This increased dependency burden makes rural households further reliant on remittances. Surprisingly, few researchers have examined remittance behavior in this new context of social and economic instability. We may expect coinsurance and investment arrangements that motivated migrants’ exchange behavior in earlier decades to be of even greater importance today as migrants seek to secure lasting assistance from rural families in the event of premature illness, death, and a legacy of dependents.
Although much progress has been made in understanding how the nature and strength of ongoing ties to the family determine remittance behavior, prior research suffers from two main limitations. First, it focuses solely on ties between migrants and their families while ignoring reciprocal relationships formed in the destination outside the origin social network. Second, perhaps because of foundations in microeconomics, current models of remittance behavior privilege material or economic resources, such as educational investments, income, and inheritable assets, as the objects of exchange. Other forms of support that are not overtly pecuniary in nature, such as psychosocial support, are rarely factored into migrants’ exchange relationships.4 The role of noneconomic forms of assistance from nonfamilial ties has been an important feature of sociological research on the adaptation process, however. According to this work, psychosocial support is crucial to migrants’ ability to deal with the initial stresses of transition and to ensure longer-term integration into the new environment (Tienda 1980; Boyd 1989; Portes 1998; Korinek, Entwisle, and Jampaklay 2005). There is a large body of research that examines the repercussions of a lack of psychosocial support in the destination, such as its negative impact on migrant physical and mental health (e.g., Kuo and Tsai 1986; Landale and Oropesa 2001; Weigers and Sherraden 2001; Barnes and Aguilar 2007; Ek et al. 2008). This work also underscores that migrants’ strategies to substitute for supportive familial relationships include recreating close bonds outside the family, kinship, or ethnic group of origin once they reach the destination (Magdol and Bessel 2003; Korinek et al. 2005; Livingston 2006). These new, nonfamilial ties and accompanying resource requirements in the destination could seriously conflict with migrants’ existing commitments to the rural family.
Transactional Sex Research in Africa
A separate body of research has been concerned with intimate relationships formed in the city, namely, nonmarital sexual partnerships, and their connection to the rapid spread of HIV/AIDS throughout sub-Saharan Africa. Emerging work on “transactional sex” examines the extent to which money and gifts are given in premarital and extramarital relationships and the consequences of these exchanges for sexual health. A number of studies have documented a high prevalence of nonmarital sexual relationships and material exchange, particularly in urban and industrialized settings (Orubuloye, Caldwell, and Caldwell 1992; Meekers and Calves 1997b; Castle and Konate 2003; Luke 2003, 2006; Dunkle et al. 2004; Kaufman and Stavrou 2004; Moore, Biddlecom, and Zulu 2007).5 These revelations have led to claims that partnerships outside marriage and the practice of transactional sex within them continue to fuel the epidemic across the continent, particularly among women (Orubuloye et al. 1992; Meekers and Calves 1997a; Côtè et al. 2004; Dunkle et al. 2004).
Much of the research on transactional sex has been conducted by researchers and policy makers who aim to locate risk factors for HIV infection and therefore focuses on sexual activities as the objects of exchange. This work generally conceives of monetary or material components of nonmarital sexual partnerships as commodity exchange, where money and gifts are assumed to be direct payments for (unsafe) sex (Meekers and Calves 1997b; Côtè et al. 2004; Dunkle et al. 2004; Chatterji et al. 2005). Numerous social scientists have criticized this view of African sexuality as promiscuous in general and commercial in particular, arguing that it reflects a Western preoccupation with morality and the commoditization of human attributes, including labor and sexuality (Kopytoff 1989; Zelizer 1994; Heald 1995; Setel 1999; Arnfred 2004; Khamasi and Maina-Chinkuyu 2005; Smith 2007). They explain that material exchange has traditionally been part of intimate sexual, marital, and familial reciprocal relationships in Africa (Bloch 1989; Bloch and Parry 1989; Helle-Valle 2005; Shipton 2007) and that transactional sex in the contemporary era must be understood as rooted in persistent gender and economic inequalities, particularly the absence of formal labor market opportunities for women. These inequalities are the sources of women’s increased risk, making it difficult for them to insist on safe sexual practices with their male partners (Hunter 2002; Luke 2003; Cole 2004; Haram 2005; Shipton 2007; Smith 2007; Swidler and Watkins 2007; Luke et al. 2009; see also Radin 2001).
This broader understanding of the social organization of African sexuality also emphasizes the diversity of nonmarital relationships and the variation in motivations and meanings for individuals who engage in them (Setel 1999; Hunter 2002; Wojcicki 2002; Cole 2004; Haram 2005; Smith 2007; Swidler and Watkins 2007). While pure commodity exchange with commercial sex workers (CSWs) or casual partners (who are often driven by financial concerns) takes place across Africa as elsewhere, a collection of qualitative studies has provided an in-depth look at the sexual relationships that men sustain with serious, longer-term girlfriends outside the marital sphere. These serious relationships more closely embody gift exchange, where transfers are imbued with particular meaning and the expectation of continued gift giving and receiving reaffirms and extends the social tie (Mauss 1990; Carrier 1991; Luke 2005b; Shipton 2007). In Kisumu, the site of my study, men refer to these partners by the emic term jadiya, which is slang for “my dear” or “my lover.” Studies from across the continent have described the range of nonmaterial benefits men receive from serious partners in addition to sexual gratification, including emotional support, romantic attachment, companionship, and care giving, and these “comforts of home” can be quite valuable to them (Dinan 1983; White 1990; Orubuloye et al. 1992; Karanja 1994; Mann 1994; Cornwall 2002; Hunter 2002; Cole 2004; Helle-Valle 2005; Smith 2007; Swidler and Watkins 2007). Indeed, serious nonmarital relationships often continue for a number of years, with men’s ongoing material gifts signaling the strength and significance of these intimate ties (Haram 2005; Maganja et al. 2007; Poulin 2007).
THEORETICAL FRAMEWORK
Both the research on remittance behavior and transactional sex envision migrants’ family networks in the origin and new ties in the destination as separate spheres of exchange, involving independent meanings, objects of exchange, and resource commitments. I argue, in contrast, that individuals are situated in a web of reciprocal relationships that often offer overlapping benefits, and they must continually make choices about how to allocate their scarce resources within and between them (Furman and Buhrmester 1985; Uehara 1990; Wellman and Wortley 1990; Zelizer 2005). I formulate a theoretical framework of competing commitments to understand how individuals negotiate these conflicting claims and predict where, why, and to what extent they make trade-offs across relationships. Although relationships provide a variety of types of benefits, my particular interest lies in the role of material commitments within familial relationships and how nonfamilial ties compete for a share of these resources.
The starting point for my framework is the notion that comparable types of social and economic support can be obtained from more than one source, and thus one relationship can substitute in function or meaning for another. Substitution is also a matter of degree. Each relationship offers a collection of benefits, and any or all of these could be duplicated by an alternative relationship. Therefore, one relationship could serve as a perfect, partial, or no substitute for another.
The concept of substitution is central to numerous theoretical perspectives within sociology. For example, social exchange theorists examine how new relationships that offer a “comparison level of alternatives” can affect satisfaction within and commitment to existing relationships (McDonald 1981; Molm 1991; Sprecher 1992; van de Rijt and Macy 2006). This work generally compares new and existing relationships within the same sphere, where the collection of benefits overlaps to a large extent, such as marital unions threatened by other sexual partners (see McDonald 1981; Felmlee, Sprecher, and Bassin 1990; Sprecher 1998). In this case, the new relationship is often a perfect substitute, and increasing engagement with the new tie leads to termination of the existing relationship. I argue that substitution can also take place across separate spheres of social interaction. New ties that offer equivalent forms of support are substitutable, regardless of their form and the extent of duplication.
Social psychologists have examined the potential for substitution across diverse spheres of intimacy. According to the “compensation model,” deficiencies in support from close familial relationships, such as parents and siblings, can be compensated by support obtained from relationships outside the family, such as friends or peers (East and Rook 1992; Gauze et al. 1996; Magdol and Bessel 2003; Sherman et al. 2006; Voorpostel and van der Lippe 2007; see also Wellman and Wortley 1990). This work is interested in the impact of substitution on individuals’ psychological adjustment—paralleling the migrant adaptation literature—which may explain its emphasis on replacement of psychosocial support of the family, including elements such as companionship, affection, trust, and emotional encouragement. In a complement to this literature, social exchange theorists have also begun to conceive of emotions as valued resources that can be the objects or outcomes of exchange between intimates (Lawler and Thye 1999).
Both the social exchange and social psychological perspectives help to explain where and why substitution occurs by identifying comparability in the benefits provided by multiple intimate relationships. Neither perspective explicitly examines resource trade-offs that occur as a consequence of substitution, however. In particular, the compensation model brings us only halfway; it focuses solely on the exchange of support across two relationships and fails to account for the give and take of resources that occur within each of them as a result of the duplication of social, economic, or psychosocial support. Social exchange theory goes a step further with its emphasis on reciprocity and the conflict produced in the existing relationship by the threat of an attractive alternative. Nevertheless, this research does not consider the important implications for resource commitments both between and within competing relationships.
To incorporate the consequences of substitution, my theoretical framework draws on the principle of substitution as conceived in microeconomic theory, which states that two goods are substitutes if they have comparable value and can be used in place of one another. Importantly, this substitutability has implications for how individuals allocate scarce resources: spending more on one good leads to spending less on the other. Both sociological and microeconomic theories recognize that substitution requires the comparability of benefits; however, the economic concept provides the missing link regarding expectations for resource trade-offs that arise as a result of this duplication.
Combining the economic principle of substitution, which is usually applied to goods and services in the marketplace, with theoretical perspectives from sociology outlined above, one can fully characterize the process of substitution across social relationships: Individuals sustain intimate relationships with a level of resource commitment that corresponds to the material and nonmaterial support received and, hence, the strength of the tie. A new relationship will substitute to the extent that it duplicates at least some of the collection of benefits offered by the existing relationship. The formation of the new relationship and its subsequent strengthening—measured by changes in resource commitments—will weaken the existing relationship with an accompanying decrease in resources allocated to it. This leads to the hypothesis that increasing transfers to the new relationship will be associated with reductions in resource commitments to the existing one, with the strength of this association depending on the degree of substitutability across the two relationships. At one extreme, engagement with a perfect substitute will lead to termination of the existing relationship and all prior resource obligations. In the case of a partial substitute, which duplicates only one of several types of support, transfers to the new relationship will decrease resource commitments and weaken the existing tie; however, resource commitments will not be curtailed for those components of support that continue to be provided by the existing relationship. At the other extreme, there will be no resource trade-offs between relationships that are not substitutable.6
The process of substitution—and the particular relationships, objects of exchange, and resources involved—depends on the social, economic, legal, and historical context (Appadurai 1989; Kopytoff 1989; Zelizer 2005; Healy 2006). In sub-Saharan Africa, the rural extended family has traditionally provided the bulk of material and nonmaterial support to its members. These close family and kinship ties have extended to the city with urbanization and occupational migration (Smith 2003; Luke and Munshi 2006), and the literature on remittance behavior finds that stronger connections to the origin are associated with larger remittances. This work neglects to recognize, however, that migrants interact with a variety of individuals and institutions in the destination, including ties acquired in the origin, such as kinsmen and wives, as well as new, nonfamilial ones established in the city. According to research on migrant adaptation, many of these ties could replicate the psychosocial support of the rural family. Among the nonfamilial ties, nonmarital sexual partners have emerged as a very important group in the current context of HIV/AIDS—not only because of their role in the spread of the disease but also because of the considerable amount of material transfers they receive.
As described above, there are multiple types of nonmarital sexual partnerships in urban Africa, which vary in form, the nature of exchange, and benefits that they provide their male partners. I expect longer-term serious partners, known as jadiya in Kisumu, to substitute for the rural extended family by providing needed psychosocial support. These are also the partners who are likely to receive the largest amounts of transfers that are sustained over the longest periods of time. To the extent that jadiya partially duplicate the support of the rural family, my theoretical framework predicts that transfers to jadiya will have a negative effect on remittances to the rural family. For commercial and casual sexual relationships that do not supply overlapping forms of support with the rural family, this negative association will be absent.
I employ the following analytical strategy to test these predictions in Kisumu. My first objective is to exploit the qualitative data to show that jadiya in the city provide not only sex but, more critically, nonmaterial support to their male partners that is similar to the psychosocial support received from the rural extended family. Second, using survey data, I test the hypothesis that a negative association should be obtained between transfers to jadiya and remittances to the family. Third, I explore additional implications of the theoretical framework—that there should be no resource trade-offs between relationships that are not substitutable in the urban African context. Finally, I consider alternative explanations for the negative jadiya transfers-remittances association. One obvious explanation is that jadiya are substitutes for migrants’ wives, and the rural family merely serves as a proxy for this duplicate relationship. We will see, however, that the negative association between transfers to jadiya and family remittances is obtained for single men as well. Another explanation is that important variables that are unobserved in the data could give rise to a spurious correlation between transfers to jadiya and remittances to the family. For example, migrants who are less attached to their rural families and remit less could also be the type of individuals who have the greatest propensity to acquire new sexual partners in the city and spend generously on them. My strategy to assess this alternative explanation exploits the idea that new relationships in the city will strengthen over time once they are formed. Thus, conditional on having a jadiya, the amount given to her should be increasing over the duration of the relationship. More important, remittances to the rural family should be decreasing over the duration of the jadiya relationship if the jadiya substitutes for the rural family. In the next section, I outline the data used in the empirical analysis before proceeding to report the series of results.
KISUMU SURVEY AND QUALITATIVE DATA
The survey and qualitative data used in the analyses were collected as part of a project I codirected in Kisumu, Kenya. Kisumu is the traditional home of the Luo ethnic group and a destination for many young migrants seeking educational and work opportunities, as well as a central town on the highway from coastal Kenya into Uganda. The high mobility and young age structure of the population are believed to have contributed to the rapid spread of HIV and other sexually transmitted diseases in this region of Kenya (Buvè et al. 2001). The researchers chose Kisumu as the site for a study of the effects of social organization on sexual behavior and labor market outcomes among a population of young adult men in an urban environment with high HIV/AIDS prevalence. Although Kisumu attracts migrants of both sexes (Lyons 2003), our project was particularly interested in examining the role of male sexual behavior in the spread of the epidemic in Western Kenya. We also chose to focus the study on the patrilineal Luo, whose migrant sons continue to be particularly important sources of economic support for rural families and communities (Hoddinott 1994).
The data are based on a random sample of Luo males ages 21–45 surveyed in 2001. Enumeration areas mapped by the government of Kenya’s Central Bureau of Statistics were used as primary sampling units within Kisumu town. Of these, 121 were randomly chosen for the survey, and all households in each enumeration area were selected. In each household, all males of eligible age were interviewed by trained field-workers. Data quality was of paramount importance to the project, particularly regarding information on sensitive behavior, such as sexual relationships and activities. Accordingly, the research team took multiple steps to ensure the validity and reliability of reporting. These measures included conducting preliminary in-depth interviews (which are discussed below) and carrying out numerous data checks in the field, such as reinterviewing 4% of the sample to confirm the reliability of responses with respect to marriage, migration, and sexual behavior (Luke 2005a). We believe that the careful attention placed on data quality resulted in a high response rate (96%) and very accurate reporting.
Remittances, Sexual Partners, and Transfers
For the purpose of studying remittance behavior, I restrict the analysis to the subsample of 2,081 men who are migrants, which I define as moving to Kisumu after birth. I also include only those who came to Kisumu at least 12 months prior to the survey, so they have had the opportunity to send remittances over the past year. The survey was designed to gather information on urban men’s ongoing commitments to rural families and communities. In Africa, perhaps more than in other societies, migrants continue to identify with their kinship groups (Geschiere and Gugler 1998; Smith 2003), and many migrants distribute their resources beyond the extended family of origin. Therefore, each respondent was first asked to estimate the total value of money and gifts in Kenyan shillings (Ksh) that he gave or sent to his rural home community in the last year, referring to his rural clan or village community. Next, each respondent was asked to estimate the total value of money and gifts that he gave or sent to his family in the last year, which refers to his household or compound of origin. Remittances to the rural home community and family were understood to be mutually exclusive and reported separately. Because my main independent variable—transfers to jadiya sexual partners—is measured in the last month, I divide the amount of family and home community remittances by 12 to obtain corresponding average monthly measures. For most of the regression specifications, the dependent variable is the total amount of remittances sent to the family in the last month. I also include one additional specification in which the value of remittances to the home community in the last month is the dependent variable.
A specific aim of the survey was to gather detailed information on male sexual behavior and material transfers to nonmarital sexual partners. Before constructing the survey instrument, the research team conducted in-depth interviews with 20 male residents of Kisumu to assist in the formulation of survey questions on the types and characteristics of nonmarital sexual partnerships in addition to questions on the marriage process. These interviews were tape-recorded and simultaneously translated and transcribed into English. For the qualitative analysis in this article, the interview data were coded and analyzed using NVivo (QSR International) to categorize the various meanings, benefits, and costs that emerged across marital and nonmarital relationships.
The survey instrument asked respondents the number of nonmarital sexual partners they had in the last year and to report information on the five most recent partners.7 Partnership information included the type of partnership (jadiya, commercial sex, or casual), the duration of the relationship in months, condom use at last sexual intercourse (yes/no), and material transfers respondents gave to each partner in the last month. A separate section focused on jadiya sexual partners, where respondents were asked if they were related to their jadiya by clan of birth, indicating that they generally stem from the same village or rural home community more broadly.8
Despite the recent surge in interest in transactional sex and money and gifts given to female sexual partners, our survey is one of the first to gather detailed information on transfers. The questions were framed to ensure that men’s reports of transfers were not explicitly tied to sexual encounters or equated with stigmatized prostitution, which could lead to underreporting (Luke 2005b). The survey question read: “It is common for men to give women gifts or other assistance when they are in a relationship. What have you given your partner in the last month?” This question was asked for each sexual partnership that took place in the last year. Response categories included the major types of transfers that were uncovered from the qualitative interviews, including money; gifts; meals, drinks, and food; rent; and an open category, where respondents could designate other items given. For each category of transfer, the respondent was asked to estimate the amount of money or value of the items given. In order to ensure accurate recall about the specific type of assistance given and the value of those transfers, the question was limited to transfers given in the last month.9 Transfers within each partnership are calculated by totaling the value of each category of assistance reported by the respondent specific to that partnership.
Control Variables
Migrants send remittances to the rural family in exchange for a range of benefits, including coinsurance against short-term economic difficulties, repayment of parental investment or to ensure future inheritance, and compensation for psychosocial support. Many migrants enter into relationships with jadiya, who provide psychosocial support in the city, and thus jadiya are a partial substitute for the rural family. According to my theoretical framework, transfers given to jadiya will reduce remittances to the family, but remittances will not be curtailed for those components of familial support that are independent of jadiya, including coinsurance and investment. The previous research on remittance behavior uses a set of variables to measure coinsurance and investment arrangements, including individual migrant and rural household characteristics. Therefore, I include this set of variables in all the regressions to control for the part of remittances that is independent of transfers to jadiya.
According to previous research, migrants’ education and income represent the migrants’ ability to remit as well as earlier parental investment in their human capital, and therefore, education and income should be positively associated with remittances (Johnson and Whitelaw 1974; Hoddinott 1994). On the survey, respondents’ self-reported income in the last year is recorded in Kenyan shillings, and I divide this value by 12 to create a monthly measure. A quadratic term is also included to allow for the possibility that the relationship between income and remittances is nonlinear. Respondents’ age is included to control for changes across the life course that may affect remittances.
Past studies have revealed that migrant marital status and residence of nuclear family members, including spouses and children, are important determinants of remittance behavior (Johnson and Whitelaw 1974; Rempel and Lobdell 1978; Menjivar et al. 1998; VanWey 2004). Migrants are concerned about the welfare of these close family members or may operate under normative expectations to provide for them; accordingly, remittances will increase if these members continue to reside in the origin and will decrease if they are living in the destination. However, if migration represents a strategy to ensure rights to inheritance and future support for these dependents, the level of remittances will be associated with the presence of these individuals in migrants’ lives, regardless of whether they reside in the origin or the destination.
These arguments focus on the residence of spouses as a key determinant of remittance behavior. In the African context, however, I believe that entry into the marital institution is an important predictor as well. The traditional system of exogamous marriage that is prevalent in much of sub-Saharan Africa (and among the Luo) has been conceptualized as a type of contractual arrangement between individuals and their kinship groups (Parkin 1978; Luke and Munshi 2006; Shipton 2007). Marriage increases the strength of ties to the patrilineage as well as the breadth of ties through the acquisition of a new affine network drawn from the kinship group of one’s spouse (Ndisi 1974; Shipton 1989; Smith 2001). The benefits of this expanded support network through marriage bring associated costs, including increased remittances to the origin family. Viewing marriage in this way, I would expect to find that currently married migrants send larger remittances than those who are single or formerly married, regardless of whether or not migrants cohabit with their spouses in the destination.
The variables for marital status are constructed from the full marital histories collected from respondents, including information on how many months of the last year they cohabited in Kisumu with their current wives.10 I construct three dummy variables designating men as currently single/unmarried (single, divorced, separated, or widowed), currently married and cohabited eight or more months of the last year with a spouse in Kisumu (which I term “spouse in destination”), and currently married and cohabited less than eight months in the last year (“spouse in origin”). The great majority of married men cohabited at least part of the year with a spouse in Kisumu; therefore, I attempt to distinguish between those whose cohabitation experience was essentially continuous from those who cohabited intermittently.11
With respect to children and their residence, respondents were asked the number of surviving children they had and how many of these children were currently living with them in Kisumu. The remaining number of children I designate as living in the origin. The data do not allow me to determine the precise household of residence for noncohabiting wives and children; however, I assume that they remain in the rural household. This assumption is quite credible for the Luo in this area of Kenya, where exogamous marriage and patrilocal residence rules oblige wives to relocate to their husbands’ family compound upon marriage, and those not cohabiting with husbands in Kisumu are very likely to be overseeing the rural homestead (Parkin 1978; Francis and Hoddinott 1993). In addition, children often remain in the rural home with their mothers or are cared for by extended family within the origin compound (Nyambedha et al. 2003).
Time and distance generally weaken migrants’ ties to the origin, which could thereby decrease remittances to the family. As migrants become established in the destination over time, they are less likely to need the family’s temporary support or coinsurance (Knowles and Anker 1981). In addition, greater distances between the origin and destination can curtail communication, visits, and the strength of familial relationships. On the contrary, if migrants and rural families maintain long-term investment arrangements, remittances should not wane over time or distance, as the future expectation of inheritance and caretaking persists. To test these claims, I include a variable indicating the number of months since the respondent migrated to Kisumu. In the sample, 86% of the migrants lived in Kisumu continuously since their first trip, and therefore, this measure provides a good approximation of migrants’ cumulative urban experience. In addition, the survey asked each respondent to identify the place of his rural home or clan community, which roughly corresponds to contemporary administrative locations, and I include a measure of the Euclidean (straight-line) distance from the center of the home location to Kisumu (White and Lindstrom 2005).
With respect to rural household characteristics, the survey collected information on the household’s wealth in the form of land and cattle, the main inheritable assets in Western Kenya (Hoddinott 1994). Investment strategies are evidenced by a positive correlation between remittances and the wealth of the rural household, which signifies the extent of potential bequests (Lucas and Stark 1985; VanWey 2004). Wealthier households also supply more risk insurance and therefore require greater remittances in return.
RESULTS
Descriptive Statistics and Qualitative Findings
In this section, I describe the sample of male Luo migrants in Kisumu and the flows of financial resources they distribute to their familial and nonfamilial ties. In addition, I use both survey and qualitative data to compare the characteristics of and transfers given to jadiya, commercial sex, and casual partners to illustrate how valued benefits and resources vary across these diverse relationships and, in particular, to confirm that jadiya provide psychosocial support in the city.
The summary statistics in table 1 describe a young and relatively educated male migrant population that is likely to be found in many urban settings across sub-Saharan Africa. In Kisumu, most migrant men were engaged in some form of employment in the last year; less than 5% of the sample was unemployed during the entire year, and almost all of these were students (not shown).12 The mean income in the last month was approximately US$70.13 The majority of migrant men are married, and the large majority of these men lived with a spouse in Kisumu for eight or more months of the last year. Most migrants came to the city as young adults and had resided in Kisumu almost nine years on average.
TABLE 1.
Summary Statistics for Migrant Men in Kisumu (N = 2,081)
| Mean or % |
SD | |
|---|---|---|
| Individual migrant characteristics: | ||
| Age (years) | 28.8 | 6.8 |
| Education (years) | 9.9 | 2.6 |
| Income in the last month (Ksh) | 4,944 | 7,103 |
| Current marital status (%): | ||
| Single (never married, divorced, separated, widowed) | 36.7 | |
| Married and spouse in origina | 15.4 | |
| Married and spouse in destinationb | 47.9 | |
| Number of children in household in Kisumu | 1.0 | 1.5 |
| Number of children outside Kisumu | .9 | 1.7 |
| Months since migrating to Kisumu | 107.6 | 83.2 |
| Rural household characteristics: | ||
| Land owned (acres) | 7.7 | 17.8 |
| Cattle owned | 5.2 | 9.2 |
| Distance to Kisumu (miles) | 28.7 | 13.2 |
Cohabited with spouse in Kisumu fewer than eight months in the last year.
Cohabited with spouse in Kisumu eight or more months in the last year.
Statistics on migrants’ resource flows are presented in table 2. The overwhelming majority of migrants (89%) remitted to their families in the last year, suggesting that remitting is indeed normative behavior in Kisumu. Among the remitters, the average amount sent to rural families in the last month (US$18) constitutes almost one-quarter of their income. Hoddinott’s (1994) study of Luo migrants in Nyanza Province found a similar likelihood and level of remitting in 1988.14 Sending remittances to the home community is also quite common among urban migrants (67%), but not as prevalent as sending resources to the family. The average amount sent was also much lower, with home community remittances making up only 5% of remitting migrants’ monthly income.
TABLE 2.
Migrants’ Resource Flows in the Last Month (N = 2,081)
| Mean or % |
SD | |
|---|---|---|
| Family: | ||
| Remitted to familya (%) | 88.6 | |
| Of those who remitted: | ||
| Amount remitted (Ksh) | 1,230 | 1,639 |
| Family remittances as a percentage of income | 22.9 | |
| Home community: | ||
| Remitted to home communitya (%) | 66.7 | |
| Of those who remitted: | ||
| Amount remitted (Ksh) | 282 | 378 |
| Home community remittances as a percentage of income | 5.0 | |
| Nonmarital sexual partners: | ||
| Had nonmarital sexual partner(s) (%) | 41.2 | |
| Of those who had nonmarital sexual partners: | ||
| Gave transfer to nonmarital sexual partner(s) (%) | 82.0 | |
| Of those who gave transfer: | ||
| Amount given to all nonmarital sexual partner(s) (Ksh) | 978 | 1,459 |
| Transfers to sexual partners as a percentage of income | 21.1 |
In the last year.
With respect to migrants’ nonfamilial ties in the form of nonmarital sexual partners, 41% of the migrants in the sample had at least one partner in the last month. Of these men, 82% gave a transfer to at least one of these partners. Like remittances, these figures indicate that giving material transfers to nonmarital sexual partners is quite common in Kisumu. Interestingly, the mean amount given over the last month (US$14) constituted over one-fifth of men’s monthly income on average and approaches the amount given to the family. In light of the view that African migrants continue to maintain close ties with the origin, the comparable figures on the amount of resources given to sexual partners suggest that the benefits migrants receive from these partnerships are quite valuable to them. Moreover, the large proportions of migrants’ incomes that are devoted to families, communities, and sexual partners suggest that competition over migrants’ resources could be quite severe.
The figures on transfers in table 2 pertain to sexual partners regardless of the nature of the relationship. A heterogeneity among nonmarital sexual partners exists in Kisumu, however, and includes jadiya, commercial sex, and casual partners. I am interested in discerning men’s motivations for engaging in these diverse relationships and the meaning of transfers within them. Analysis of the in-depth interview and survey data provides more detailed descriptions of these sexual partnership types.
The survey data indicate that jadiya and casual relationships are the most common types of nonmarital sexual partnerships among migrant men in Kisumu. Approximately one-third (31%) of the migrant men in the sample were involved with at least one jadiya sexual partner in the last month, 17% were engaged with at least one casual partner, and only 2% had encounters with CSWs. A recent decline in commercial sex has been documented in Kisumu and other urban African settings (Voeten et al. 2007), presumably the result of HIV/AIDS education and increased awareness about the risks of commercial sex. In the later stages of the epidemic in Western Kenya, it nevertheless appears customary for men to engage in nonmarital partnerships in the city, and relationships with jadiya serious girlfriends are particularly common.
The data from the in-depth interviews illuminate the nature of these nonmarital relationships in the city. It is clear from men’s descriptions that nonmarital sexual relationships embody social exchange and that this exchange is gendered, particularly with respect to the types of resources given and received. Respondents agree that male partners most often provide money and material assistance to their female partners, whereas women supply sexual gratification and nonmaterial forms of support, including affection and emotional encouragement. In all of men’s accounts, no one mentioned receiving financial or material resources from their female partners.15 These findings echo the research on mate selection across cultures, which shows that men generally exchange their wealth and social standing for women’s physical attractiveness, sexual availability, and emotional support and care (Wellman and Wortley 1990; Sprecher 1998; van de Rijt and Macy 2006).
In the in-depth interviews, men were asked to describe the difference between jadiya and casual relationships. Respondents drew a sharp distinction between the types of support received from these partners and men’s commitment to them. Casual relationships were described as short-term “hit-and-run” encounters with “passersby,” whose main purpose is for sex. They are transitory and less serious by nature and involve little emotional or personal investment. In contrast, jadiya relationships are longer-term commitments that involve much more than sexual intercourse. Jadiya are described using the language of love, intimacy, and trust. Respondents spoke of the “emotional attachment,” “emotional bond,” and “closeness” formed with these serious girlfriends, which appear to be major benefits for urban men. Indeed, many men said they hope to eventually marry these partners or continue to be involved with them as “outside wives” once they do marry since they are not able to break the emotional bond that has developed (Orubuloye et al. 1992; Fennell and Luke 2007).16
Although jadiya relationships vary in their intensity, there was considerable consensus across respondents in the meaning of these partnerships and the major forms of support that they provide. Following are several respondents’ descriptions of the differences between jadiya and casual partners:
It depends on the intimacy. There’s what you call “hit-and-run” casual. But there’s somebody where the relationship continues for a longer period, and this intimate relationship is what they call jadiya. But a casual partner is just a day-to-day hit-and-run, and there’s no relationship. Your interest is just to have sex and go away. (Married, age 39, no nonmarital sexual partners)
I think casual partners are those who cover weekends or one-night stands with a man but are never to be seen again. Others [jadiya] get more involved and are more regular and are depended on by their partners. (Married, age 33, multiple nonmarital partners)
They both provide sex all right, … but a jadiya is one who is much closer. (Married, age 22, one jadiya)
These qualitative findings are supported by the statistics describing men’s nonmarital sexual partnerships by type in table 3. The seriousness of jadiya relationships is reflected in their length, which is one and one-half years on average, much longer than the other types of partnerships. Past studies have found that partnership duration is also associated with issues of trust and condom use: as partners know each other longer, trust between them increases and condom use decreases (Harrison, Xaba, and Kunene 2002; Tavory and Swidler 2009). Among the sample of migrants, men used condoms at last sexual intercourse in only 39% of jadiya relationships, whereas condom use was substantially higher in casual and commercial partnerships. Although men were not asked about their relationships with CSWs in the in-depth interviews, I assume that the main purpose of these partnerships is sexual gratification, and they entail little personal investment, which is reflected in their short duration.
TABLE 3.
Characteristics of Migrants’ Nonmarital Sexual Partners by Type
| Mean or % |
|||
|---|---|---|---|
| Jadiya | Casual | CSWs | |
| Length of relationship (months) | 17.3 | 6.8 | 2.2 |
| Used condom at last sex (%) | 38.6 | 63.2 | 85.3 |
| Received transfer in last month (%) | 76.3 | 63.6 | 94.2 |
| Amount of transfers in last month (Ksh) (conditional on giving transfers) | 680.2 | 379.1 | 412.4 |
| Started relationship after moving to Kisumu (%) | 92.4 | 98.1 | 100.0 |
| Related to respondent (%) | 1.4 | … | … |
| N | 776 | 484 | 69 |
Note.—All differences across partner types are significant at the P < .001 level using analysis of variance.
Stark contrasts also appeared between the amount of transfers given to jadiya and casual partners and the meaning attached to these transfers. Numerous men remarked that jadiya relationships are “expensive” and that men give jadiya larger transfers than other types of nonmarital partners. As shown in table 3, over three-quarters of jadiya received a transfer in the month before the survey, and they receive much larger amounts than casual and commercial sex partners on average. CSWs are the most likely to receive a transfer; of course, these relationships are commercial by definition. The value of the compensation to CSWs is similar to the value of transfers to casual partners. The qualitative interviews reveal that transfers to jadiya also reflect a strong interdependence between partners: Men continue to give material assistance to their jadiya to ensure that these relationships—and the benefits they receive from them—continue. At the same time, men realize that jadiya rely on them for sustained financial support. In contrast, transfers to casual partners are fleeting and represent impersonal payments for sex.
The survey data also shed light on the process of relationship formation. Comparing the date of migrants’ arrival in Kisumu with the duration of their sexual partnerships reveals that nonmarital partnerships are overwhelmingly new ties men develop after moving to Kisumu. Jadiya relationships, in particular, are formed in the city with unrelated women, and thus jadiya are not extensions of the family or community who receive a share of origin remittances in Kisumu.
In sum, the descriptive statistics and qualitative data reveal that migrant men engage in nonmarital partnerships upon arriving in the city, and these new relationships embody social exchange: They require various amounts of material expenditure, and men receive various forms of support in return. The remittances literature finds a positive association between the strength of ties in the origin and the level of remittances. Nonfamilial relationships in the destination show a similar association, with larger transfers exchanged for the superior benefits of relationships with jadiya sexual partners. In particular, jadiya provide needed forms of psychosocial support, such as emotional closeness, intimacy, and trust, which I argue could replace the support of the rural extended family and thereby reduce remittances to these relations.
Jadiya as Substitutes for the Rural Family
To test my hypothesis regarding the trade-off between transfers to jadiya and remittances to the family, I present the results of ordinary least squares (OLS) regressions in table 4. The amount of remittances to the family in the last month is the dependent variable. The full sample of migrants is included, and those who did not report remitting are assigned “0” for the amount of remittances. Apart from transfers to jadiya, the regressors include individual migrant and rural household characteristics that have been used to measure coinsurance and investment arrangements with the rural family in past research.
TABLE 4.
Regression Analysis of the Determinants of Remittances to the Family (N = 2,081)
| (1) |
(2) |
|||
|---|---|---|---|---|
| Coef. | SE | Coef. | SE | |
| Sexual partners: | ||||
| Transfers to jadiya sexual partner | −.09 | .04* | −.09 | .04* |
| Has a jadiya sexual partner | … | … | −21.25 | 77.18 |
| Individual migrant characteristics: | ||||
| Age | −6.41 | 7.20 | −6.54 | 7.22 |
| Education | 8.75 | 12.55 | 8.43 | 12.61 |
| Income/1,000 | 149.13 | 8.12*** | 149.17 | 8.12*** |
| Income squared | −.52 | .04*** | −.52 | 1.04*** |
| Current marital status: | ||||
| Single (ref.) | … | … | … | … |
| Married and spouse in origina | 316.86 | 107.64** | 310.50 | 110.12** |
| Married and spouse in destinationb | 363.29 | 86.51*** | 357.23 | 89.29*** |
| Number of children in household in Kisumu | 59.52 | 30.17* | 59.61 | 30.18* |
| Number of children outside Kisumu | 83.04 | 23.75*** | 83.08 | 23.75*** |
| Months since migrating | −.02 | .40 | −.03 | .40 |
| Rural household characteristics: | ||||
| Land owned | 3.72 | 2.21+ | 3.72 | 2.21+ |
| Cattle owned | 2.12 | 3.49 | 2.11 | 3.49 |
| Distance from Kisumu | −3.50 | 2.35 | −3.50 | 2.35 |
| Constant | 217.41 | 218.56 | 234.05 | 226.81 |
Cohabited with spouse in Kisumu fewer than eight months in the last year.
Cohabited with spouse in Kisumu eight or more months in the last year.
P<.10.
P<.05.
P<.01.
P<.001.
The results in column 1 show that the amount given to jadiya sexual partners has a negative and significant effect on remittances, which supports my contention that jadiya substitute for the rural family. All men in the sample are included in the regression, whether they gave transfers to a jadiya or not. Thus, the variable for jadiya transfers is coded 0 in the absence of transfers in the last month and 0 in the absence of any jadiya relationship as well. To distinguish between these observations, the specification in column 2 includes a dummy variable for having a jadiya to estimate the effect of transfers to jadiya conditional on having one as a sexual partner. The transfers coefficient is essentially unchanged from column 1 and remains significant, whereas the dummy variable is insignificant, suggesting that having a jadiya by itself has no direct effect on family remittances. It is the amount that is given to her that decreases the amount of resources sent to the family. With respect to the magnitude of the transfers effect, we see in table 3 that the average transfer to a jadiya in the last month is Ksh 680; the estimated coefficients in table 4 imply that engaging in a jadiya relationship and giving the average amount of transfers to her reduces remittances by Ksh 83, or 8% of monthly remittances on average. This is not a trivial amount and may represent sizable reductions for rural families reliant on this assistance.17
I find several interesting results with respect to migrants’ individual characteristics and remittance behavior in table 4. While the linear effect of income is positive and significant, the coefficient on the quadratic term is negative (and significant), implying that the income effect is declining at the margin. A comparison of the linear and quadratic coefficients nevertheless shows that remittances are increasing with income over the entire range of incomes reported in my sample. Migrants remit smaller amounts at low levels of income, possibly because they have yet to adjust to urban life and establish steady employment at that point in their careers (Cliggett 2003). Although remittances do increase monotonically with income thereafter, the declining marginal effect could reflect an investment arrangement, where migrants with higher earnings gain less from potential parental bequests (Hoddinott 1994).
Marital status is an important determinant of remittance behavior in Kisumu. Currently single men (the reference category) send significantly smaller amounts to their families than men with wives in the destination and men with wives in the origin. A separate analysis shows that there is no significant difference between the two categories of married men (not shown). Taken together, these results indicate that it is not the residence of the spouse that matters for family remittances, but the marriage institution more broadly. This supports the previous assertion that exogamous marriage—like migration—is a social exchange arrangement between men and their families that is accompanied by increased financial obligations. Remittances also increase significantly with each additional child living within the household in Kisumu and also with each child residing outside of Kisumu, whom I presume to be living in the origin location. Both effects suggest an investment strategy, but the latter (significantly larger) effect also suggests compensation to rural families for the care they provide these dependents.
With respect to the rural household’s characteristics, land owned displays a positive and marginally significant association with the level of material resources remitted, providing evidence of an investment arrangement. Finally, it is important to note that years of residence in Kisumu and the distance the rural household lies from Kisumu display insignificant associations with the amount of remittances. Overall, it appears that neither time nor distance significantly weakens African migrants’ ties to the origin (Guilmoto 1998).
On the whole, the results in table 2 and table 4 reveal several noteworthy conclusions. I find that migrants maintain reciprocal arrangements with rural familial and urban nonfamilial ties concurrently, demonstrating migrants’ position in the “dual system” in urban Africa. The overwhelming majority of male migrants continue to remit to their families, and the high levels appear to be sustained when compared to estimates of remittance behavior several decades ago. The hypothesized relationships for a long-term investment arrangement with the rural household hold up most consistently, as seen in the significant effects of income, wives, children, and land and the absence of effects of time and distance; the absence of a time effect also reveals that coinsurance is a not a primary motivation. Investment arrangements were advantageous to African migrants and rural families according to earlier studies; however, the motivations behind current connections may be linked to repercussions of the ongoing HIV/AIDS epidemic in the region.
The series of results also support the argument that resource trade-offs between migrants’ familial and nonfamilial ties will depend on the degree of substitutability with the rural family. Jadiya in the city substitute for nonmaterial support, and thus migrants’ resources allocated to the family are partially curtailed. In particular, the qualitative evidence suggests that the nature of this support is psychosocial, including aspects such as emotional attachment and trust. The rural family continues to provide migrants with valuable economic benefits that jadiya in the city do not, and therefore, migrants maintain their relationships with the family and continue remitting a share of their resources. It appears that jadiya sexual partners are important, but not perfect, substitutes for the rural family of origin.
Additional Implications and Alternative Explanations
The qualitative and quantitative results presented thus far support the argument that substitution between competing commitments accounts for the trade-off between familial and nonfamilial ties in Kisumu and that transfers to jadiya lead to reductions in remittances to the rural family. In this subsection, I provide further support for the theoretical framework by first considering additional implications and, second, by exploring alternative explanations for the negative transfers-remittances association.
The main additional implication of my theoretical framework is that resource trade-offs will not take place between relationships that are not substitutes. To test this prediction, I examine trade-offs between different types of sexual partners in the city and various types of remittances to the rural origin. As we have seen, the major benefit of commercial sex and casual partnerships is sexual gratification, which does not duplicate the support provided by the origin extended family. Therefore, I do not expect to observe a negative and significant association between transfers to commercial sex or casual partners and family remittances. The benefits for migrants of relationships with the home community, such as social status and patron-client connections (Smith 2003, 2007), are unlikely to be provided to migrants by jadiya sexual partners.18 In this case as well, I do not expect to find a negative and significant association between transfers to jadiya and remittances to the home community.
I examine these possibilities by conducting OLS regressions of the determinants of family and home community remittances. The results are shown in table 5. Each specification includes the same set of individual migrant and rural household variables as table 4 to control for existing family ties. Because commercial sex and casual partnerships are similar in their characteristics and a small number of men are involved with CSWs, I pool transfers to commercial sex and casual partners in the analysis.
TABLE 5.
Regression Analysis of the Determinants of Remittances to the Family and Home Community(N = 2,081)
| Family Remittances (1) |
Home Community Remittances (2) |
Family Remittances (3) |
||||
|---|---|---|---|---|---|---|
| Coeff. | SE | Coef. | SE | Coef. | SE | |
| Sexual partners: | ||||||
| Transfers to commercial sex/casual partner | .03 | .07 | … | … | … | … |
| Transfers to jadiya sexual partner | … | … | .02 | .01 + | −.08 | .05 + |
| Individual migrant characteristics: | ||||||
| Age | −6.07 | 7.22 | 3.56 | 1.61* | −6.45 | 7.20 |
| Education | 7.15 | 12.55 | 7.25 | 2.81* | 8.94 | 12.57 |
| Income/1,000 | 146.53 | 8.12*** | 22.15 | 1.82*** | 149.16 | 8.13*** |
| Income squared | −.51 | .04*** | −.08 | .01*** | −.52 | .04*** |
| Current marital status: Single (reference) | … | … | … | … | … | … |
| Married and spouse in origina | 349.14 | 106.91** | 45.35 | 24.12 + | 330.78 | 112.00** |
| Married and spouse in destinationb | 397.57 | 85.43*** | 48.06 | 19.38* | 374.79 | 89.57*** |
| Married/spouse in origin × transfers to jadiya | … | … | … | … | −.05 | .14 |
| Married/spouse in destination × transfers to jadiya | … | … | … | … | −.03 | .10 |
| Number of children in household in Kisumu | 61.53 | 30.19* | −1.50 | 6.76 | 59.31 | 30.23* |
| Number of children outside Kisumu | 82.49 | 23.78** | 3.99 | 5.32 | 82.95 | 23.76*** |
| Months since migrating | −.04 | .40 | −.07 | .09 | −.03 | .40 |
| Rural household characteristics: | ||||||
| Land owned | 3.71 | 2.21 + | .03 | .50 | 3.68 | 2.21 + |
| Cattle owned | .79 | 3.47 | 2.45 | 7g** | 2.05 | 3.50 |
| Distance from Kisumu | −3.80 | 2.35 | −.01 | .53 | −3.48 | 2.35 |
| Constant | 200.05 | 218.96 | −132.08 | 48.98** | 211.21 | 219.10 |
Cohabited with spouse in Kisumu fewer than eight months in the last year.
Cohabited with spouse in Kisumu eight or more months in the last year.
P<.10.
P<.05.
P<.01.
P<.001.
We see in column 1 that the association between transfers to commercial sex/casual partners and remittances to the family is small, positive, and insignificant. In column 2, transfers to jadiya have a small, positive, and marginally significant association with home community remittances. As expected, I fail to observe a negative and significant association between either of these types of transfers and remittances. These findings provide support for my theoretical framework and the notion that resource tradeoffs occur only between relationships that are substitutable. The findings also rule out an alternative explanation for the estimated negative jadiya transfers-family remittances association—that, with a fixed budget, any new financial outlay will reduce the amount of resources migrants have available to distribute to all existing relationships.19
It is also interesting to note that a positive and significant association between jadiya transfers and home community remittances is observed. This finding could be interpreted in two ways. First, in contrast to the potential for relationships to serve as substitutes, microeconomic theory allows for the possibility that two goods could be complements. Here, the benefits of a new relationship supplement the benefits of the existing relationship, and thus expenditures devoted to the new relationship would increase commitments to the existing one. In this context, jadiya relationships are unlikely to be complementary to relations with the rural clan community, however. The second and more likely explanation is that the positive correlation between jadiya transfers and home community remittances is due to a selection effect, where unobserved characteristics of migrants, such as generosity or extravagance, lead to greater expenditures on jadiya as well as the kinship group. I discuss additional consequences of such unobserved heterogeneity in greater detail below.
To preserve space, I do not discuss the coefficients on individual and household characteristics in table 5 but note that the coefficients in column 1 are generally similar to the results for family remittances in table 4. The determinants of home community remittances in column 2 differ from those of the family; in particular, the positive effects of age, education, and income suggest that contemporary African migrants maintain these ties to secure status and the instrumental benefits of wider kinship connections in the origin.
Thus far, I have argued that jadiya serious partners in the city act as substitutes for the traditional sphere of the rural family. Another obvious alternative explanation for the negative association between jadiya transfers and family remittances is that substitution stems from partners of the same type and that jadiya are substituting for wives in particular, not extended families. This alternative explanation echoes research on extramarital and extradyadic sex undertaken by social exchange theorists in the United States (Sprecher 1998). We may expect wives to provide the same forms of psychosocial support and sexual gratification as serious nonmarital sexual partners. Thus, the negative transfers-remittances association found in earlier regressions could essentially be picking up the trade-off in resources between jadiya and wives among the married men in the sample.
I examine this alternative by estimating the effect of transfers to jadiya on family remittances for single men in table 5, column 3. The same regression specification is used as in table 4, with interaction terms included between jadiya transfers and migrants’ marital/cohabitation status (single men, married men with wives in the origin, and married men with wives in the destination). With single men as the reference category, the coefficient on jadiya transfers is interpreted as the effect for single men only; this coefficient is negative, marginally significant, and of similar magnitude to the coefficient in table 4 for the full sample. This finding upholds the view that jadiya substitute for the support offered by the rural extended family and not merely for wives. The negative coefficients on the interaction terms for married men, although insignificant, suggest a greater degree of substitution for them, perhaps a trade-off between jadiya and wives in addition to that with rural families.
On the whole, the results in table 5, column 3, underscore the important distinctions between extended family, marital, and serious extramarital relationships in contemporary Western Kenya. For example, it appears that essential psychosocial support continues to be provided by the rural extended family and has not become the major domain of spouses. This finding supports scholars who argue that lineage ties are stronger and closer than marital ones in many parts of Africa (Caldwell, Caldwell, and Quiggen 1989; Adepoju and Mbugua 1997; Dodoo and Frost 2008). In addition, the finding that resource trade-offs are largely unaffected by the presence of a wife or her location of residence suggests that jadiya are not perfect or even close substitutes for wives, contrary to the role of extramarital partners in other settings. Indeed, past work has shown that although African marriage is evolving to incorporate more aspects of companionate unions, many men continue to find intimacy and emotional support outside of marriage rather than within it (Karanja 1994; Smith 2001, 2007; Cole 2004; Haram 2005; Helle-Valle 2005; Fennell and Luke 2007).20
A final, alternative explanation for the negative association between family remittances and jadiya transfers that I consider is unobserved heterogeneity; that is, migrants who are less connected to the origin are also the type of individuals who are more likely to form new nonmarital relationships in the city. The weak positive association between jadiya transfers and remittances to the home community in table 5, column 2, rules out the possibility that individuals who are generally less attached to their origin location (both the family and home community) give more to their jadiya partners. Conversely, the absence of a negative and significant association between transfers to commercial sex and casual partners and remittances to the family in table 5, column 1, rules out the possibility that individuals who give more to all their nonmarital partners in the city give less to their rural families. If there is an unobserved individual characteristic that is generating a spurious negative correlation, it must be specific to the jadiya-family relationship, which is a more stringent requirement.
The regressions in table 4 and table 5 include an extensive list of characteristics that have been previously documented to determine remittances. In an attempt to rule out the possibility that unobserved migrant characteristics nevertheless continue to independently determine both remittances and jadiya transfers, I introduce a new variable in the analysis— the length of the jadiya relationship—that determines the level of transfers for any individual. The longer one maintains a jadiya relationship, the stronger the tie should be, and the greater the level of transfers; however, there is no a priori reason to believe that the length of the jadiya relationship in the city should directly determine the level of remittances to the rural family except through its effect on jadiya transfers. Thus, if I do find a negative association between the length of the jadiya relationship and remittances, this would provide strong support for the trade-off between jadiya and the family.
The results of the analysis using length of the jadiya relationship as a determinant of both transfers and family remittances are presented in table 6. Column 1 shows that, conditional on having a jadiya, the length of the relationship has a positive and highly significant association with the level of transfers, as expected, signaling a strengthening of this intimate tie over time. Column 2 shows that, conditional on having a jadiya, the length of the relationship has a negative and marginally significant association with remittances to the family, supporting the view that a gradual strengthening of the jadiya relationship over time, with its accompanying increase in transfers, leads to a decline in remittances. The coefficients on the individual and rural household characteristics that determine remittances in column 2 are similar to the corresponding coefficients in the original specification (table 4, col. 2). In addition, looking at column 1 and the determinants of jadiya transfers, we see, not surprisingly, that education, income, and rural household wealth increase these resource flows. Moreover, married migrants with spouses in Kisumu give less to jadiya than other men, perhaps because of their greater resource commitments to wives in the city or because wives directly oversee resource allocation.
TABLE 6.
Regression Analysis of the Determinants of Transfers toJadiya Sexual Partners and Remittances to the Family (N = 2,081)
| Transfers to Jadiya (1) |
Family Remittances (2) |
Family Remittances (TV) (3) |
||||
|---|---|---|---|---|---|---|
| Coef. | SE | Coef. | SE | Coef. | SE | |
| Sexual partners: | ||||||
| Length of jadiya relationship (months) | 7.12 | .96*** | −3.25 | 1.91 + | … | … |
| Transfers to jadiya sexual partner | … | … | … | … | −.46 | .27 + |
| Has a jadiya sexual partner | 498.27 | 38.71*** | −32.23 | 76.66 | 195.47 | 176.06 |
| Individual migrant characteristics: | ||||||
| Age | 3.91 | 3.65 | −7.51 | 7.23 | −5.72 | 7.37 |
| Education | 18.02 | 6.39** | 8.86 | 12.66 | 17.10 | 14.30 |
| Income/1,000 | 20.31 | 4 08*** | 147.44 | 8.08*** | 156.73 | 9.92*** |
| Income squared | −.09 | 02*** | −.51 | 04*** | −.55 | .05*** |
| Current marital status: Single (reference) | … | … | … | … | ||
| Married and spouse in origina | −90.69 | 55.80 | 304.13 | 110.49** | 262.69 | 117.34* |
| Married and spouse in destinationb | −113.93 | 45.32* | 352.06 | 89 74*** | 300.00 | 99.95** |
| Number of children in household in Kisumu | −23.63 | 15.24 | 62.98 | 30.19* | 52.18 | 31.18+ |
| Number of children outside Kisumu | −4.04 | 12.02 | 85.48 | 23.80*** | 83.63 | 24.18** |
| Months since migrating | .17 | .20 | −.02 | .40 | .06 | .41 |
| Rural household characteristics: | ||||||
| Land owned | .46 | 1.12 | 3.61 | 2.21 | 3.82 | 2.25 + |
| Cattle owned | 11.68 | 1.75*** | 1.17 | 3.46 | 6.50 | 4.78 |
| Distance from Kisumu | 2.29 | 1.19+ | −3.64 | 2.35 | −2.59 | 2.48 |
| Constant | −376.82 | 114.50** | 283.33 | 226.73 | 111.13 | 247.53 |
Cohabited with spouse in Kisumu fewer than eight months in the last year.
Cohabited with spouse in Kisumu eight or more months in the last year.
P<.10.
P<.05.
P<.01.
P<.001.
An alternative way of thinking about this analysis is that the length of the jadiya relationship serves as a statistical instrument for transfers to jadiya (Moffitt 2005). Column 1 in table 6 would then correspond to the first-stage regression, showing that the instrument significantly affects transfers to jadiya, and column 2 would be the reduced-form regression. The corresponding instrumental variable (IV) estimates are presented in column 3, which give us consistent estimates of the causal effect of jadiya transfers on family remittances if the conditions for a valid instrument are satisfied. We see that the jadiya transfers coefficient continues to be (marginally) significant and is larger in magnitude than the coefficient in table 4, column 2 (although the coefficient on the jadiya dummy variable increases as well). The effect of engaging in a jadiya relationship and providing the average amount of transfers to her results in a reduction of 11% of migrants’ monthly remittances on average, which is larger than what was obtained in the OLS regression in table 4.
In order for an instrument to be valid, it must satisfy the exclusion restriction. In this case, the identifying assumption is that, conditional on being in a jadiya relationship and controlling for all the variables usually assumed to determine family remittances in the literature, the length of the jadiya relationship should not have a direct effect on remittances to the family. The basic idea underlying my strategy is that while particular types of individuals may choose to acquire a jadiya once they arrive in Kisumu, the delay before a partnership materializes and, hence, the observed length of the relationship are plausibly uncorrelated with characteristics that directly determine remittances. Although this identifying assumption cannot be tested directly (Moffitt 2005), note that the jadiya dummy is insignificant in columns 2 and 3 of table 6 as well as in column 2 of table 4. The decision to enter into a jadiya relationship appears to be uncorrelated with characteristics that directly determine remittances (although the empirical strategy allows for this possibility). We would thus expect the subsequent delay in acquiring a jadiya, among those who made that choice, to be uncorrelated with these characteristics as well, satisfying the requirements for a valid instrument.
It is also worth noting that the length of the jadiya relationship has no effect on transfers to the home community and, conversely, that the length of a relationship with commercial sex/casual partners has no effect on family remittances (not shown). These results further emphasize the special relationship between transfers to jadiya and remittances to the rural extended family. Finally, notice that the length of time in Kisumu has no effect on family remittances, and hence jadiya transfers, in table 6, columns 1 and 2. Without the “shock” of a particular type of nonmarital relationship in the city (the jadiya), migrants appear to maintain the strength of their ties to the family over time, suggesting that the direction of the relationship appears to be jadiya transfers leading to reductions in family remittances, not the reverse.21
The analysis in this section tests the hypothesis that entry into a new relationship, with its accompanying transfers, should be associated with a decline in resource commitments to existing substitute relationships. As predicted by the theoretical framework, I find that increased transfers to new jadiya sexual relationships in the city are associated with a statistically significant reduction in remittances to the rural extended family. In contrast, a negative and significant association is not observed between resources provided to relationships that are not substitutes. To investigate the robustness of these results, I carry out a series of independent tests to rule out alternative explanations for the observed negative association between jadiya transfers and family remittances. First, I rule out the possibility that the negative association is a mechanical consequence of resource allocations being made subject to a fixed budget constraint. Second, I provide evidence that this relationship is not driven by an underlying substitution between jadiya and wives. Finally, I provide evidence that reduces the possibility that unobserved individual characteristics independently determine remittances and jadiya transfers, giving rise to a spurious negative correlation. While statistical inference with cross-sectional data does not provide proof of causality, it is encouraging that all these independent tests support the basic hypothesis—that there is indeed substitution across separate spheres of exchange in Kisumu.
CONCLUSION
Sociologists recognize that individuals engage in a web of reciprocal relationships involving various forms of support and accompanying resource obligations. When faced with commitments to multiple exchange partners, individuals must make choices about how to allocate their scarce resources within and between them. I develop a theoretical framework to explain how individuals negotiate these conflicting claims and predict where, why, and to what extent they make trade-offs across relationships. My framework draws on multiple theoretical perspectives, which contend that relationships substitute for one another if they provide comparable forms of social and economic support. My contribution to this work is to emphasize the consequences of substitution for resource allocation between competing relationships. The theoretical framework predicts that the increase in resources devoted to a new relationship will be associated with a decline in resources committed to an existing relationship as long as the two are substitutes. The strength of this negative association is determined by the degree of substitution. When there is no substitutability across relationships, no trade-off in resources will be observed.
The theory of substitution can be applied to a range of social settings involving diverse relationships, objects of exchange, and resource commitments. I test my theoretical framework among urban migrants in contemporary Kisumu, Kenya. In this context, the rural extended family provides migrants with valuable material and nonmaterial benefits, and migrants send large amounts of remittances to these rural relations in return. At the same time, migrants create new ties and devote large financial sums to serious nonmarital sexual partners in the city, who duplicate the family’s nonmaterial, psychosocial support and thereby compete for a share of migrants’ resources. As my theoretical framework predicts, transfers allocated to serious sexual partners significantly decrease remittances to the family, by more than 10% on average. Additional tests find that this negative and significant association is absent among relationships in the origin and destination that are not substitutes. In a context characterized by high HIV/AIDS and continuing economic uncertainty, the trade-offs that I uncover have important implications for the stability of many families in rural Kenya. Given the scale of remittances migrants send globally, competing commitments in the destination could potentially affect resources distributed to origin families worldwide.
My work has focused on two relationships that have been identified by separate literatures as being particularly relevant in urban Africa: The remittances literature focuses on the significance of migrants’ ties to the family of origin, whereas the work on transactional sex highlights men’s nonmarital sexual partnerships in the city. However, other relationships formed in the city could substitute for the rural family as well. For example, an increase in the strength of religious organizations in response to the economic and emotional fallout of the AIDS epidemic has been documented in many African settings (Adogame 2007; Agadjanian and Sen 2007; Dageid and Duckert 2008). An interesting and important area for future research is the nature of competition between institutions, such as the modern church and the traditional family, and the potential for substitution and resource trade-offs that arise as a consequence.
The current study focuses on the choices that migrant men make between competing relationships; in future work, it would be useful to shift the perspective to their female partners and investigate their motivations for entering into nonmarital relationships. Existing research throughout sub-Saharan Africa finds that women seek serious, committed, and trusting relationships, which are accompanied by material benefits and possibly marriage (e.g., Dinan 1983; Cole 2004). Other work emphasizes the instrumental nature of nonmarital relationships. Many women satisfy their partners’ need for psychosocial and sexual support in return for valuable material resources, often from multiple partners concurrently or sequentially, with marriage an unlikely prospect (e.g., Meekers and Calves 1997b; Silberschmidt and Rasch 2001; Wojcicki 2002; Luke 2003; Hunter 2007). In this case, women’s sexual and psychosocial “work” are compensated and commoditized to some degree (Hochschild 1979). While numerous traditional familial responsibilities, such as child care and elder care, are available for purchase from market institutions in industrialized societies (Folbre 2001; Hochschild 2003), it has been argued that the singular nature of sentimental and psychosocial support makes them more resistant to duplication and hence commoditization (Appadurai 1989; Kopytoff 1989; Folbre 2008). It is consequently interesting to observe the possible emergence of a market for psychosocial support in Africa, where markets for insurance, credit, formal employment, and health services have yet to fully develop. Further exploration of the emergence of markets in both social and economic spheres of life and the forms of support that are substituted and potentially commoditized is another important avenue for future research.
Footnotes
This article is based on data from a project directed by Nancy Luke and Kaivan Munshi, Population Studies and Training Center, Brown University. I gratefully acknowledge support from the World Bank; the National Institute on Aging (grant AG12836) through the Population Aging Research Center at the University of Pennsylvania; and the Mellon Foundation, the Center for AIDS Research, and the University Research Foundation at the University of Pennsylvania. I thank Francis Ayuka and the Survey Research Team, Nairobi, for their superb work collecting the data. Nitsan Chorev, David Lindstrom, Bruce Meyer, Kaivan Munshi, Martin Piotrowski, Holly Reed, Susan Short, and Michael White as well as seminar participants at the University of Chicago, Princeton University, Boston University, and Brown University provided helpful comments on earlier drafts. I also thank Julie Fennell for qualitative research assistance and Kelley Smith for editorial assistance.
Most studies conceive of remittances as resource flows to households of origin, although the theoretical literature commonly refers to migrants’ relationships with families of origin as well. Although I recognize that families and households are not synonymous within the social science literature, I use these terms interchangeably in this article.
The area in and around Kisumu did not have low-cost, easy access to antiretroviral treatment at the time of the study in 2001 and continues to have limited access today.
Nonmaterial forms of support have been variously defined and analyzed in the sociological and psychological literature. I am particularly interested in the general category of psychosocial support, which includes such elements as companionship, enhancement of self-worth, affection, trust, closeness, and emotional support (see Wellman and Wortley 1990; East and Rook 1992; Gauze et al. 1996; Magdol and Bessel 2003; Sherman, Lansford, and Volling 2006; Voorpostel and van der Lippe 2007).
Although females may give material transfers to their male partners, most studies find that transfers flow more often and in vastly larger amounts from men to their female partners (Meekers and Calves 1997b; Matasha et al. 1998; Castle and Konate 2003; Luke et al. 2009).
Numerous researchers within the social exchange tradition have also considered how commitment, solidarity, and affective regard influence interactions within a given relationship once confronted with attractive alternatives (e.g., Lawler and Yoon 1996; Molm, Schaefer, and Collett 2007). On the whole, these sentiments “heighten one’s comparison level for alternatives” (McDonald 1981, p. 833) and thereby could diminish the strength of the observed association between resource commitments to new and existing ties. However, we would still expect to see a negative association after taking these other influences into account.
Of the men reporting nonmarital sexual partners in the last year, 97.5% had five or fewer partners.
Exogamous marriage rules prohibit Luos from marrying a relative, evena very distant one, and we wanted to investigate if relationships with serious girlfriends also adhered to this prescription.
Given that transfers to sexual partners are measured in the last month, two types of measurement error could arise. First, classical measurement error occurs if there is random variation of transfers across months. In this case, the estimated coefficient on transfers would be biased toward zero. Second, there could be a seasonal aspect to transfers, where men are more able to give transfers during certain times of the year or women may demand more during specific months, since the survey was conducted at one point in time. This type of error could bias the coefficient on transfers in either direction. The instrumental variable estimates reported in table 6, col. 3, below correct for both sources of measurement error.
As in many African societies, marriage among the Luo involves numerous events and negotiations, which may take place over a few months to two years (Southall 1973). On the survey, marriage was marked by the time a bride moves into her husband’s home (ikendo), which is generally regarded as the official marriage date (Ndisi 1974; Ocholla-Ayayo 1976). Using this emic term and further detailed questions on marriage and jadiya relationships on the survey, we also believe that respondents clearly distinguished actual marriages from serious nonmarital relationships with jadiya.
Of the married men, 6.2% are currently polygynous and may cohabit in Kisumu with more than one wife at a time or consecutively. For the variable pertaining to the number of months of cohabitation, I add the number of months these respondents cohabited with each wife; those who cohabited more than 12 months (2% of respondents) are recoded as living with a spouse 12 months in the last year.
These and all subsequent results not shown are available from the author on request.
Seventy Kenyan shillings were the equivalent of US$1.00 at the time of the survey.
Johnson and Whitelaw’s (1974) study of urban migrants in Nairobi in 1971 found a similar prevalence of remitting as well.
Other studies find that serious female partners also provide cooking and other domestic services for their male partners (Luke 2003; Smith 2007). Migrant men may have obtained this support from some jadiya in Kisumu as well; however, these benefits may not have been as important as psychosocial support and therefore may not have been spontaneously mentioned in the interviews.
While some men eventually marry their jadiya, most jadiya are unlikely to become wives. For example, approximately 30% of married men in Kisumu had a jadiya in the last year. If each of them married these partners, the urban polygyny rate would increase from 6.2% of married men to over 30%, a figure that would be excessively large for contemporary urban Africa (Lesthaeghe, Kaufmann, and Meekers 1989; Timaeus and Reynar 1998).
One question is whether migrants could offer other forms of support, such as increased visits or communication, in lieu of material remittances to the rural family. Nonpecuniary support, regardless of its form, is unlikely to be a perfect substitute for material remittances in this context, however. Rural families enter into exchange relationships with their migrant sons to gain valued material remittances, and they cannot be fully compensated with other resources that are less beneficial to them.
Men in urban Kisumu as well as other African cities underscore the importance of keeping extramarital partners secret from the family and wider community (Longfield et al. 2004; Fennell and Luke 2007; Smith 2007), and thus they do not confer status within these groups. Attracting extramarital partners may display sexual and economic prowess among men’s more restricted peer groups, however (Smith 2007).
The results in tables 4 and 5 also rule out a normative explanation for migrants’ remittance behavior, which holds that social norms and role expectations guide these actions (Uehara 1990). A normative perspective expects migrants to continue to remit to their rural families and home communities and at fixed levels regardless of new ties with sexual partners and the amount of money and gifts that are given to them.
The qualitative and survey data also support the view that the role of jadiya is distinct from that of wives, and therefore, many men maintain both types of relationships. In-depth interview participants explain that jadiya relationships are primarily sustained by emotional and romantic attachment, whereas marital relationships are sustained by familial and lineage obligations. Furthermore, cohabitation and child-bearing are expected behaviors among married couples, and the survey data show that 76% of married men cohabited with a spouse in the last year in Kisumu and 87% have a child with their wives; corresponding figures among all men in jadiya relationships are less than 1% and 10%, respectively.
While I presume that new relationships weaken existing ties and resources devoted to them, it could be the case that the deterioration of existing relationships (and accompanying resources) compels individuals to search for new substitutes and hence provide greater resources to them. Either scenario produces the same trade-off according to the theoretical framework; the direction of substitution is, however, specific to the context. In Kisumu, the evidence just presented suggests that ongoing strong ties to the rural extended family are weakened by the strengthening of new relationships with jadiya
REFERENCES
- Adepoju Aderanti, Mbugua Wariara. “The African Family: An Overview of Changing Forms.”. In: Adepoju Aderanti., editor. Family, Population and Development in Africa. London: Zed Books; 1997. pp. 42–57. [Google Scholar]
- Adogame Afe. “HIV/AIDS Support and African Pentecostalism.”. Journal of Health Psychology. 2007;12(3):475–484. doi: 10.1177/1359105307076234. [DOI] [PubMed] [Google Scholar]
- Agadjanian Victor, Sen Soma. “Promises and Challenges of Faith-Based AIDS Care and Support in Mozambique.”. American Journal of Public Health. 2007;97(2):362–366. doi: 10.2105/AJPH.2006.085662. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Appadurai Arjun. “Introduction: Commodities and the Politics of Value.”. In: Appadurai Arjun., editor. The Social Life of Things. Cambridge: Cambridge University Press; 1989. pp. 3–63. [Google Scholar]
- Arnfred Signe. “Re-thinking Sexualities in Africa: Introduction.”. In: Arnfred Signe., editor. Re-thinking Sexualities in Africa. Uppsala: Almqvist & Wiksell Tryckeri AB; 2004. pp. 7–29. [Google Scholar]
- Azam Jean-Paul, Gubert Flore. “Migrants’ Remittances and the Household in Africa: A Review of Evidence.”. Journal of African Economies. 2006;15(suppl. 2):426–462. [Google Scholar]
- Barnes Donelle M, Aguilar Rocio. “Community Social Support for Cuban Refugees in Texas.”. Qualitative Health Research. 2007;17(2):225–237. doi: 10.1177/1049732306297756. [DOI] [PubMed] [Google Scholar]
- Bloch Maurice. “The Symbolism of Money in Imerina.”. In: Parry Jonathan, Bloch Maurice., editors. Money and the Morality of Exchange. Cambridge: Cambridge University Press; 1989. pp. 16–190. [Google Scholar]
- Bloch Maurice, Parry Jonathan. “Introduction: Money and the Morality of Exchange.”. In: Parry Jonathan, Bloch Maurice., editors. Money and the Morality of Exchange. Cambridge: Cambridge University Press; 1989. pp. 1–32. [Google Scholar]
- Boyd Monica. “Family and Personal Networks in International Migration: Recent Developments and New Agendas.”. International Migration Review. 1989;23(3):638–670. [PubMed] [Google Scholar]
- Brockerhoff Martin, Biddlecom Ann E. “Migration, Sexual Behavior, and the Risk of HIV Infection in Kenya.”. International Migration Review. 1999;33(4):833–856. [Google Scholar]
- Buvè A, Lagarde E, Carael M, Rutenberg N, Ferry B, Glynn JR, Laourou M, Akam E, Chege J, Sukwa T for the Study Group on Heterogeneity of HIV Epidemics in African Cities. “The Epidemiology of Gonorrhoea, Chlamydial Infection and Syphilis in Four African Cities.”. AIDS. 2001;15(suppl. 4):S79–S88. doi: 10.1097/00002030-200108004-00009. [DOI] [PubMed] [Google Scholar]
- Caldwell John C, Anarfi John K, Caldwell Pat. “Mobility, Migration, Sex, STDs, and AIDS: An Essay on Sub-Saharan Africa with Other Parallels.”. In: Herdt Gilbert., editor. Sexual Cultures and Migration in the Era of AIDS. Oxford: Clarendon; 1997. pp. 41–54. [Google Scholar]
- Caldwell John C, Caldwell Pat, Quiggen Pat. “The Social Context of AIDS in Sub-Saharan Africa.”. Population and Development Review. 1989;15(2):185–234. [Google Scholar]
- Carrier James. “Gifts, Commodities, and Social Relations: A Maussian View of Exchange.”. Sociological Forum. 1991;6(1):119–136. [Google Scholar]
- Castle Sarah, Konate Mamadou Kani. “Economic Transactions Associated with Sexual Intercourse among Malian Adolescents: Implications for Sexual Health.”. In: Agyei-Mensah Samuel, Casterline John B., editors. Reproduction and Social Context in Sub-Saharan Africa. Westport, Conn.: Greenwood; 2003. pp. 159–186. [Google Scholar]
- Chatterji Minki, Murray Nancy, Longon D, Anglewicz Philip. “The Factors Influencing Transactional Sex among Young Men and Women in 12 Sub-Saharan African Countries.”. Social Biology. 2005;52(1–2):56–72. doi: 10.1080/19485565.2002.9989099. [DOI] [PubMed] [Google Scholar]
- Chimwaza AF, Watkins SC. “Giving Care to People with Symptoms of AIDS in Rural Sub-Saharan Africa.”. AIDS Care. 2004;16(7):795–807. doi: 10.1080/09540120412331290211. [DOI] [PubMed] [Google Scholar]
- Clark Samuel J, Collinson Mark A, Kahn Kathleen, Drullinger Kyle, Tollman Stephen M. “Returning Home to Die: Circular Labor Migration and Mortality in South Africa.”. Scandinavian Journal of Public Health. 2007;35(suppl. 69):35–44. doi: 10.1080/14034950701355619. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Cliggett Lisa. “Gift Remitting and Alliance Building in Zambian Modernity: Old Answers to Modern Problems.”. American Anthropologist. 2003;105(3):1–10. [Google Scholar]
- Cole Jennifer. “Fresh Contact in Tamatave, Madagascar: Sex, Money, and Intergenerational Transformation.”. American Ethnologist. 2004;31(4):573–588. [Google Scholar]
- Cornwall Andrea. “Spending Power: Love, Money, and the Reconfiguration of Gender Relations in Ado-Odo, Southwestern Nigeria.”. American Ethnologist. 2002;29(4):963–980. [Google Scholar]
- Côtè Anne-Marie, Sobela Francois, Dzukoto Agnes, Nzambi Khonde, Asamoah-Adu Comfort, Labbè Annie-Claude, Masse Benoit, Mensah Joyce, Frost Eric, Pepin Jacques. “Transactional Sex Is the Driving Force in the Dynamics of HIV in Accra, Ghana.”. AIDS. 2004;18:917–925. doi: 10.1097/00002030-200404090-00009. [DOI] [PubMed] [Google Scholar]
- Dageid Wenche, Duckert Fanny. “Balancing between Normality and Social Death: Black, Rural, South African Women Coping with HIV/AIDS.”. Qualitative Health Research. 2008;18(2):182–195. doi: 10.1177/1049732307312070. [DOI] [PubMed] [Google Scholar]
- de la Brière Bénédicte, Sadoulet Elisabeth, Janvry Alain de, Lambert Sylvie. “The Roles of Destination, Gender, and Household Composition in Explaining Remittances: An Analysis for the Dominican Sierra.”. Journal of Development Economics. 2002;68:309–328. [Google Scholar]
- Dinan Carmel. “Sugar Daddies and Gold-Diggers: The White-Collar Single Women in Accra.”. In: Oppong Christine., editor. Female and Male in West Africa. London: Allen & Unwin; 1983. pp. 344–366. [Google Scholar]
- Dodoo F. Nii-Ammoo, Frost Ashley E. “Gender in African Population Research: The Fertility/Reproductive Health Example.”. Annual Review of Sociology. 2008;34:431–452. [Google Scholar]
- Dunkle Kristin L, Jewkes Rachel K, Brown Heather C, Gray Glenda E, McIntryre James A, Harlow Sioban D. “Transactional Sex among Women in Soweto, South Africa: Prevalence, Risk Factors and Association with HIV Infection.”. Social Science and Medicine. 2004;59:1581–1592. doi: 10.1016/j.socscimed.2004.02.003. [DOI] [PubMed] [Google Scholar]
- East Patricia L, Rook Karen S. “Compensatory Patterns of Support among Children’s Peer Relationships: A Test Using School Friends, Nonschool Friends, and Siblings.”. Developmental Psychology. 1992;28(1):163–172. doi: 10.1037/0012-1649.28.1.163. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Ek Ellen, Koiranen Markku, Raatikka Veli-Pekka, Jarvelin Marjo-Riitta, Taanila Anja. “Psychosocial Factors as Mediators between Migration and Subjective Well-Being among Young Finnish Adults.”. Social Science and Medicine. 2008;66:1545–1556. doi: 10.1016/j.socscimed.2007.12.018. [DOI] [PubMed] [Google Scholar]
- Felmlee Diana, Sprecher Susan, Bassin Edward. “The Dissolution of Intimate Relationships: A Hazard Model.”. Social Psychology Quarterly. 1990;53(1):13–30. [Google Scholar]
- Fennell Julie, Luke Nancy. “Wives and ‘Jadiya’ Lovers: Parallel Unions in Post-polygynous Kenya.”. Manuscript: Brown University; 2007. [Google Scholar]
- Findley Sally E. “Migration and Family Interactions in Africa.”. In: Adepoju Aderanti., editor. Family, Population and Development in Africa. London: Zed Books; 1997. pp. 107–138. [Google Scholar]
- Folbre Nancy. The Invisible Heart. New York: New Press; 2001. [Google Scholar]
- Folbre Nancy. “When a Commodity Is Not Exactly a Commodity.”. Science. 2008;319:1769–1770. doi: 10.1126/science.1153904. [DOI] [PubMed] [Google Scholar]
- Francis Elizabeth, Hoddinott John. “Migration and Differentiation in Western Kenya: A Tale of Two Sub-locations.”. Journal of Development Studies. 1993;30(1):115–145. [Google Scholar]
- Furman Wyndol, Buhrmester Duane. “Children’s Perceptions of the Personal Relationships in Their Social Networks.”. Developmental Psychology. 1985;21(6):1016–1024. [Google Scholar]
- Gauze Cyma, Bukowski William M, Aquan-Assee Jasmin, Sippola Lorrie K. “Interactions between Family Environment and Friendship and Associations with Self-Perceived Well-Being during Early Adolescence.”. Child Development. 1996;67:2201–2216. [PubMed] [Google Scholar]
- Geschiere Peter, Gugler Joseph. “The Urban-Rural Connection: Changing Issues of Belonging and Identification.”. Africa. 1998;68(3):309–319. [Google Scholar]
- Glynn JR, Carael M, Auvert B, Kahindo M, Chege J, Musonda R, Kaona F, Buve A and the Study Group on the Heterogeneity of HIV Epidemics in African Cities. “Why Do Young Women Have a Much Higher Prevalence of HIV than Young Men? A Study in Kisumu, Kenya and Ndola, Zambia.”. AIDS. 2001;1515(suppl. 4):S51–S60. doi: 10.1097/00002030-200108004-00006. [DOI] [PubMed] [Google Scholar]
- Gugler Joseph. “Life in a Dual System Revisited: Urban-Rural Ties in Enugu, Nigeria, 1961-87.”. World Development. 1991;19(5):399–409. [Google Scholar]
- Guilmoto Christophe Z. “Institution and Migrations: Short-Term versus Long-Term Moves in Rural West Africa.”. Population Studies. 1998;52(1):85–103. [Google Scholar]
- Haram Liv. “‘Prostitutes’ or Modern Women? Negotiating Respectability in Northern Tanzania.”. In: Arnfred Signe., editor. Re-thinking Sexualities in Africa. Uppsala: Almqvist & Wiksell Tryckeri AB; 2005. pp. 211–229. [Google Scholar]
- Harrison Abigail, Xaba Nonhlanhla, Kunene Pinky. “Understanding Safe Sex: Gender Narratives of HIV and Pregnancy Prevention by Rural South African School-Going Youth.”. Reproductive Health Matters. 2002;9(17):63–71. doi: 10.1016/s0968-8080(01)90009-6. [DOI] [PubMed] [Google Scholar]
- Heald Suzette. “The Power of Sex: Some Reflections on the Caldwells’ ‘African Sexuality’ Thesis.”. Africa. 1995;65(4):489–505. [PubMed] [Google Scholar]
- Healy Kieran. Last Best Gifts. Chicago: University of Chicago Press; 2006. [Google Scholar]
- Helle-Valle Jo. “Understanding Sexuality in Africa: Diversity and Contextualized Individuality.”. In: Arnfred Signe., editor. Re-thinking Sexualities in Africa. Uppsala: Almqvist & Wiksell Tryckeri AB; 2005. [Google Scholar]
- Hochschild Arlie Russell. “Emotion Work, Feeling Rules, and Social Structure.”. American Journal of Sociology. 1979;85(3):551–575. [Google Scholar]
- Hochschild Arlie Russell. The Commercialization of Intimate Life: Notes from Home and Work. Berkeley and Los Angeles: University of California Press; 2003. [Google Scholar]
- Hoddinott John. “A Model of Migration and Remittances Applied to Western Kenya.”. Oxford Economic Papers. 1994;46:459–476. [Google Scholar]
- Hosegood Victoria, Preston-Whyte Eleanor, Busza Joanna, Moitse Sindile, Timaeus Ian M. “Revealing the Full Extent of Households’ Experiences of HIV and AIDS in Rural South Africa.”. Social Science and Medicine. 2007;65:1249–1259. doi: 10.1016/j.socscimed.2007.05.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Hunter Mark. “The Materiality of Everyday Sex: Thinking beyond ‘Prostitution.’”. African Studies. 2002;61(1):99–120. [Google Scholar]
- Hunter Mark. “The Changing Political Economy of Sex in South Africa: The Significance of Unemployment and Inequalities to the Scale of the AIDS Pandemic.”. Social Science and Medicine. 2007;64:689–700. doi: 10.1016/j.socscimed.2006.09.015. [DOI] [PubMed] [Google Scholar]
- Itzigsohn Jose. “Migrant Remittances, Labor Markets, and Household Strategies: A Comparative Analysis of Low-Income Household Strategies in the Caribbean Basin.”. Social Forces. 1995;74(2):633–655. [Google Scholar]
- Johnson GE, Whitelaw WE. “Urban-Rural Income Transfers in Kenya: An Estimated-Remittances Function.”. Economic Development and Cultural Change. 1974;22:473–479. [Google Scholar]
- Karanja Wambui Wa. “The Phenomenon of ‘Outside Wives’: Some Reflections on Its Possible Influence on Fertility.”. In: Bledsoe C, Pison G, editors. Nuptiality in Sub-Saharan Africa. Oxford: Clarendon; 1994. pp. 194–214. [Google Scholar]
- Kaufman Carol E, Stavrou Stavros E. “‘Bus Fare Please’: The Economics of Sex and Gifts among Young People in Urban South Africa.”. Cultural, Health and Sexuality. 2004;6(5):377–391. [Google Scholar]
- Khamasi Jennifer Wanjiku, Maina-Chinkuyu Susan Nyambura., editors. Sexuality: An African Perspective. Eloret, Kenya: Moi University Press; 2005. [Google Scholar]
- Knowles James C, Anker Richard. “An Analysis of Income Transfers in a Developing Country: The Case of Kenya.”. Journal of Development Economics. 1981;8:205–226. [Google Scholar]
- Kopytoff Igor. “The Cultural Biography of Things: Commoditization as Process.”. In: Appadurai Arjun., editor. The Social Life of Things. Cambridge: Cambridge University Press; 1989. pp. 64–91. [Google Scholar]
- Korinek Kim, Entwisle Barbara, Jampaklay Aree. “Through Thick and Thin: Layers of Social Ties and Urban Settlement among Thai Migrants.”. American Sociological Review. 2005;70:779–800. [Google Scholar]
- Kuo Wen H, Tsai Yung-Mei. “Social Networking, Hardiness and Immigrant’s Mental Health.”. Journal of Health and Social Behavior. 1986;27:133–149. [PubMed] [Google Scholar]
- Landale Nancy S, Oropesa RS. “Migration, Social Support and Perinatal Health: An Origin-Destination Analysis of Puerto Rican Women.”. Journal of Health and Social Behavior. 2001;42:166–183. [PubMed] [Google Scholar]
- Lawler Edward J, Thye Shane R. “Bringing Emotions into Social Exchange Theory.”. Annual Review of Sociology. 1999;25:217–244. [Google Scholar]
- Lawler Edward J, Yoon Jeongkoo. “Commitment in Exchange Relations: Test of a Theory of Relational Cohesion.”. American Sociological Review. 1996;61:89–108. [Google Scholar]
- Lesthaeghe RJ, Kaufmann G, Meekers D. “The Nuptiality Regimes in Sub-Saharan Africa.”. In: Lesthaeghe RJ, editor. Reproduction and Social Organization in Sub-Saharan Africa. Berkeley and Los Angeles: University of California Press; 1989. pp. 238–333. [Google Scholar]
- Livingston Gretchen. “Gender, Job Searching, and Employment Outcomes among Mexican Immigrants.”. Population Research and Policy Review. 2006;25:43–66. [Google Scholar]
- Longfield Kim, Glick Anne, Waithaka Margaret, Berman John. “Relationships between Older Men and Younger Women: Implications for STIs/HIV in Kenya.”. Studies in Family Planning. 2004;35(2):125–134. doi: 10.1111/j.1728-4465.2004.00014.x. [DOI] [PubMed] [Google Scholar]
- Lucas Robert EB, Stark Oded. “Motivations to Remit: Evidence from Botswana.”. Journal of Political Economy. 1985;93(5):901–918. [Google Scholar]
- Luke Nancy. “Age and Economic Asymmetries in the Sexual Relationships of Adolescent Girls in Sub-Saharan Africa.”. Studies in Family Planning. 2003;34(2):67–86. doi: 10.1111/j.1728-4465.2003.00067.x. [DOI] [PubMed] [Google Scholar]
- Luke Nancy. “Confronting the ‘Sugar Daddy’ Stereotype: Age and Economic Asymmetries and Risky Sexual Behavior in Urban Kenya.”. International Family Planning Perspectives. 2005a;31(1):6–14. doi: 10.1363/3100605. [DOI] [PubMed] [Google Scholar]
- Luke Nancy. “Investigating Exchange in Sexual Relationships in Sub-Saharan Africa Using Survey Data.”. In: Jejeebhoy Shireen J, Shah Iqbal, Thapa Shyam., editors. Sex without Consent: Young People in Developing Countries. London: Zed Books; 2005b. pp. 105–124. [Google Scholar]
- Luke Nancy. “Exchange and Condom Use in Informal Sexual Relationships in Urban Kenya.”. Economic Development and Cultural Change. 2006;54(2):319–348. [Google Scholar]
- Luke Nancy, Goldberg Rachel, Mberu Blessing, Zulu Eliya. “Employment, Transactional Sex, and Bargaining within Premarital Relationships in Urban Kenya.”. Manuscript: Brown University; 2009. [Google Scholar]
- Luke Nancy, Munshi Kaivan. “New Roles for Marriage in Urban Africa: Kinship Networks and the Labor Market in Kenya.”. Review of Economics and Statistics. 2006;88(2):264–282. [Google Scholar]
- Lurie Mark N, Williams Brian G, Zuma Khangelani, Mkaya-Mwamburi David, Garnett Geoff P, Sturm Adriaan W, Sweat Michael D, Gittelsohn Joel, Abdool Karim Salim S. “The Impact of Migration on HIV-1 Transmission in South Africa: A Study of Migrant and Non-migrant Men and Their Partners.”. Sexually Transmitted Diseases. 2003;30(2):149–156. doi: 10.1097/00007435-200302000-00011. [DOI] [PubMed] [Google Scholar]
- Lyons Maryinez. “Mobile Populations and HIV/AIDS in East Africa.”. In: Kalipeni Ezekiel, Craddock Susan, Oppong Joseph R, Ghosh Jayati., editors. HIV and AIDS in Africa: Beyond Epidemiology. Malden, Mass: Blackwell; 2003. pp. 175–182. [Google Scholar]
- Maganja RK, Maman S, Groues A, Mbwambo JK. “Skinning the Goat and Pulling the Load: Transactional Sex among Youth in Dar es Salaam, Tanzania.”. AIDS Care. 2007;19(8):974–981. doi: 10.1080/09540120701294286. [DOI] [PubMed] [Google Scholar]
- Magdol Lynn, Bessel Diane R. “Social Capital, Social Currency, and Portable Assets: The Impact of Residential Mobility on Exchanges of Social Support.”. Personal Relationships. 2003;10:149–169. [Google Scholar]
- Mann Kristin. “The Historical Roots and Cultural Logic of Outside Marriage in Colonial Lagos.”. In: Bledsoe C, Pison G, editors. Nuptiality in Sub-Saharan Africa: Contemporary Anthropological and Demographic Perspectives. Oxford: Clarendon; 1994. pp. 167–193. [Google Scholar]
- Matasha E, Ntembelea T, Payaud P, Saidi W, Todd J, Mujaya B, Tendo-Wambua L. “Sexual and Reproductive Health among Primary and Secondary School Pupils in Mwanza, Tanzania: Need for Intervention.”. AIDS Care. 1998;10(5):571–582. doi: 10.1080/09540129848433. [DOI] [PubMed] [Google Scholar]
- Mauss Marcel. The Gift. New York: Norton; 1990. [Google Scholar]
- McDonald Gerald W. “Structural Exchange and Marital Interaction.”. Journal of Marriage and the Family. 1981;43(4):825–839. [Google Scholar]
- Meekers Dominique, Calves Anne-Emmanuele. Washington, D.C.: PSI Research Division; 1997a. “Gender Differentials in Adolescent Sexual Activity and Reproductive Health Risks in Cameroon”. Working Paper no. 4. [Google Scholar]
- Meekers Dominique, Calves Anne-Emmanuele. “‘Main’ Girlfriends, Girlfriends, Marriage, and Money: The Social Context of HIV Risk Behavior in Sub-Saharan Africa.”. Health Transition Review. 1997b;7(suppl.):361–375. [PubMed] [Google Scholar]
- Menjivar Cecilia, DaVanzo Julie, Greenwell Lisa, Valdez R Burciaga. “Remittance Behavior among Salvadoran and Filipino Immigrants in Los Angeles.”. International Migration Review. 1998;32(1):97–126. [Google Scholar]
- Moffitt Robert. “Remarks on the Analysis of Causal Relationships in Population Research.”. Demography. 2005;42(1):91–108. doi: 10.1353/dem.2005.0006. [DOI] [PubMed] [Google Scholar]
- Molm Linda D. “Affect and Social Exchange: Satisfaction in Power-Dependence Relations.”. American Sociological Review. 1991;56(4):475–493. [Google Scholar]
- Molm Linda D, Schaefer David R, Collett Jessica L. “The Value of Reciprocity.”. Social Psychology Quarterly. 2007;70(2):199–217. [Google Scholar]
- Mooney Margarita. “Migrants’ Social Ties in the U.S. and Investment in Mexico.”. Social Forces. 2003;81(3):1147–1170. [Google Scholar]
- Moore Ann M, Biddlecom Ann E, Zulu Eliya M. “Prevalence and Meanings of Exchange of Money and Gifts for Sex in Unmarried Adolescent Sexual Relationships in Sub-Saharan Africa.”. African Journal of Reproductive Health. 2007;11(3):44–61. [PMC free article] [PubMed] [Google Scholar]
- NASCOP (National AIDS and STI Control Programme) AIDS in Kenya: Trends, Interventions, Impact. 7th ed. Nairobi: Ministry of Health; 2005. [Google Scholar]
- Ndisi John W. A Study in the Economic and Social Life of the Luo of Kenya. Lund: Berlingska Boktryckeriet; 1974. [Google Scholar]
- Nyambedha Erick Otieno, Wandibba Simiyu, Aagaard-Hansen Jens. “Changing Patterns of Orphan Care Due to the HIV Epidemic in Western Kenya.”. Social Science and Medicine. 2003;57:301–311. doi: 10.1016/s0277-9536(02)00359-3. [DOI] [PubMed] [Google Scholar]
- Ocholla-Ayayo ABC. Traditional Ideology and Ethics among the Southern Luo. Uppsala: Scandinavian Institute of African Studies; 1976. [Google Scholar]
- Orubuloye IO, Caldwell John C, Caldwell Pat. “Diffusion and Focus in Sexual Networking: Identifying Partners and Partners’ Partners.”. Studies in Family Planning. 1992;23(6):343–351. [PubMed] [Google Scholar]
- Parker Erin M, Short Susan E. “Grandmother Co-residence, Maternal Orphans, and School Enrollment in Sub-Saharan Africa.”. Journal of Family Issues. 2009;30(6):813–836. doi: 10.1177/0192513X09331921. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Parkin David. The Cultural Definition of Political Response: Lineal Destiny among the Luo. London: Academic Press; 1978. [Google Scholar]
- Piotrowski Martin. “The Effect of Social Networks at Origin Communities on Migrant Remittances: Evidence from Nang Rong District.”. European Journal of Population. 2006;22:67–94. [Google Scholar]
- Portes Alejandro. “Social Capital: Its Origins and Applications in Modern Sociology.”. Annual Review of Sociology. 1998;24:1–24. [Google Scholar]
- Poulin Michelle. “Sex, Money, and Premarital Partnerships in Southern Malawi.”. Social Science and Medicine. 2007;65:2383–2393. doi: 10.1016/j.socscimed.2007.05.030. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Radin Margaret Jane. Contested Commodities: The Trouble with Trade in Sex, Children, Body Parts, and Other Things. Cambridge, Mass: Harvard University Press; 2001. [Google Scholar]
- Rempel Henry, Lobdell Richard A. “The Role of Urban-to-Rural Remittance in Rural Development.”. Journal of Development Studies. 1978;14:324–341. [Google Scholar]
- Sana Mariano. “Buying Membership in the Transnational Community: Migrant Remittances, Social Status, and Assimilation.”. Population Research and Policy Review. 2005;24:231–261. [Google Scholar]
- Schatz Enid, Ogunmefun Catherine. “Caring and Contributing: The Role of Older Women in Rural South African Multi-generational Households in the HIV/ AIDS Era.”. World Development. 2007;35(8):1390–1403. [Google Scholar]
- Setel Philip W. A Plague of Paradoxes. Chicago: University of Chicago Press; 1999. [Google Scholar]
- Sherman Aurora M, Lansford Jennifer E, Volling Brenda L. “Sibling Relationships and Best Friendships in Young Adulthood: Warmth, Conflict, and Well-Being.”. Personal Relationships. 2006;13:151–165. [Google Scholar]
- Shipton Parker. Bitter Money: Cultural Economy and Some African Meanings of Forbidden Commodities. American Ethnological Society Monograph Series no. 1. Washington, D.C.: American Anthropological Association; 1989. [Google Scholar]
- Shipton Parker. The Nature of Entrustment. New Haven, Conn: Yale University Press; 2007. [Google Scholar]
- Silberschmidt Margrethe, Rasch Vibeke. “Adolescent Girls, Illegal Abortions and ‘Sugar Daddies’ in Dar es Salaam: Vulnerable Victim and Active Social Agents.”. Social Science and Medicine. 2001;52:1815–1826. doi: 10.1016/s0277-9536(00)00299-9. [DOI] [PubMed] [Google Scholar]
- Smith Daniel Jordan. “Romance, Parenthood, and Gender in a Modern African Society.”. Ethnology. 2001;40(2):129–151. [Google Scholar]
- Smith Daniel Jordan. “Patronage, Per Diems, and the ‘Workshop Mentality’: The Practice of Family Planning Programs in Southeastern Nigeria.”. World Development. 2003;31(4):703–715. [Google Scholar]
- Smith Daniel Jordan. “Modern Marriage, Men’s Extramarital Sex, and HIV Risk in Southeastern Nigeria.”. American Journal of Public Health. 2007;97(6):997–1005. doi: 10.2105/AJPH.2006.088583. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Southall Aidan W. “The Luo of South Nyanza, Western Kenya.”. In: Molnos A, editor. Cultural Source Materials for Population Planning in East Africa. vol. 3. Nairobi: East African Publishing House; 1973. [Google Scholar]
- Sprecher Susan. “How Men and Women Expect to Feel and Behave in Response to Inequity in Close Relationships.”. Social Psychology Quarterly. 1992;55(1):57–69. [Google Scholar]
- Sprecher Susan. “Social Exchange Theories and Sexuality.”. Journal of Sex Research. 1998;35:32–43. [Google Scholar]
- Stark Oded, Lucas Robert EB. “Migration, Remittances, and the Family.”. Economic Development and Cultural Change. 1988;36(3):465–481. [Google Scholar]
- Swidler Ann, Watkins Susan Cotts. “Ties of Dependence: AIDS and Transactional Sex in Rural Malawi.”. Studies in Family Planning. 2007;38(3):147–162. doi: 10.1111/j.1728-4465.2007.00127.x. [DOI] [PubMed] [Google Scholar]
- Tavory Iddo, Swidler Ann. “Condom Semiotics: Meaning and Condom Use in Rural Malawi.”. American Sociological Review. 2009;74:171–189. [Google Scholar]
- Tienda Marta. “Familialism and Structural Assimilation of Mexican Immigrants in the United States.”. International Migration Review. 1980;14(3):383–408. [PubMed] [Google Scholar]
- Timaeus Ian M, Reynar Angela. “Polygynists and Their Wives in Sub-Saharan Africa: An Analysis of Five Demographic and Health Surveys.”. Population Studies. 1998;52:145–162. [Google Scholar]
- Uehara Edwina. “Dual Exchange Theory, Social Networks, and Informal Social Support.”. American Journal of Sociology. 1990;96(3):521–557. [Google Scholar]
- van de Rijt Arnout, Macy Michael W. “Power and Dependence in Intimate Exchange.”. Social Forces. 2006;84(3):1455–1470. [Google Scholar]
- VanWey Leah K. “Altruistic and Contractual Remittances between Male and Female Migrants and Households in Rural Thailand.”. Demography. 2004;41(4):739–756. doi: 10.1353/dem.2004.0039. [DOI] [PubMed] [Google Scholar]
- Voeten HACM, Egesha OB, Varkevisser CM, Habbema JDF. “Female Sex Workers and Unsafe Sex in Urban and Rural Nyanza, Kenya: Regular Partners May Contribute More to HIV Transmission than Clients.”. Tropical Medicine and International Health. 2007;12(2):174–182. doi: 10.1111/j.1365-3156.2006.01776.x. [DOI] [PubMed] [Google Scholar]
- Voorpostel Marieke, van der Lippe Tanja. “Support between Siblings and between Friends: Two Worlds Apart?”. Journal of Marriage and Family. 2007;69:1271–1282. [Google Scholar]
- Weigers Margaret E, Sherraden Marget S. “A Critical Examination of Acculturation: The Impact of Health Behaviors, Social Support and Economic Resources on Birth Weight among Women of Mexican Descent.”. International Migration Review. 2001;35(3):804–839. [Google Scholar]
- Wellman Barry, Wortley Scot. “Different Strokes from Different Folks: Community Ties and Social Support.”. American Journal of Sociology. 1990;96(3):558–588. [Google Scholar]
- White Louise. The Comforts of Home: Prostitution in Colonial Nairobi. Chicago: University of Chicago Press; 1990. [Google Scholar]
- White Michael J, Lindstrom David P. “Internal Migration.”. In: Poston Dudley L, Micklin Michael., editors. Handbook of Population. New York: Kluwer; 2005. pp. 311–346. [Google Scholar]
- Wojcicki Janet M. “Commercial Sex Work or Ukuphanda? Sex-for-Money Exchange in Soweto and Hammanskraal Area, South Africa.”. Culture, Medicine, and Psychiatry. 2002;26(3):339–370. doi: 10.1023/a:1021291922026. [DOI] [PubMed] [Google Scholar]
- Zelizer Viviana A. Pricing the Priceless Child. Princeton, N.J.: Princeton University Press; 1994. [Google Scholar]
- Zelizer Viviana A. The Purchase of Intimacy. Princeton, N.J.: Princeton University Press; 2005. [Google Scholar]