Cooperative childrearing1 is a central feature of the human cooperative niche, supporting mothers2 and their dependent offspring (henceforth, children, where children can refer to biological, adopted, and/or stepchildren; Hrdy, 2007, Hrdy, 2009). Though the literature has traditionally focused on the role of kin and affinal kin in cooperative childrearing, researchers are increasingly recognizing the flexibility in who provides care (e.g., Crittenden & Marlowe, 2008; Emmott, Page, & Myers, 2020; E. Emmott & Page, 2019; Gottlieb, 2009; Myers, Page, & Emmott, 2021). Some have documented the role of non-kin, like neighbors and older non-kin children, in providing direct care or indirect care to the child – that is, via the mother – with impacts including calories available to children and mothers (Jang, Janmaat, Kandza, & Boyette, 2022), higher survival rates for children (Borgerhoff Mulder & Beheim, 2011), and increased fertility and fecundity for mothers (Page et al., 2021).

Despite increasing evidence of non-kin contributions to cooperative childrearing, explicit investigation of non-kin allomaternal care remains rare (but see Crittenden & Marlowe, 2008; Page et al., 2019)—meaning that while we have hypotheses about why non-kin provide care, and the extent to which non-kin inputs translate into benefits for mothers and children, the data to test these ideas are few. Here, we specifically focus on the benefits non-kin may provide to children by examining the impacts of godparents on child educational outcomes in two rural Bolivian communities. We find that having godparents does not predict educational attainment among adults; however, having more older siblings does.

Studies of extant forager groups suggest that human groups have always included non-kin members. For each individual band member in recent surveys of the Ju/’hoansi and Aché, for example, between 25 and 50% of band members were unrelated or distantly related individuals (Hill et al., 2011); in other words, non-kin were often available as allomothers, where allomaternal refers to care provided by those other than the mother. Research with contemporary foragers (Crittenden & Marlowe, 2008; Meehan, 2009; Page et al., 2019), horticulturalists and subsistence farmers (Gottlieb, 2009; Meehan, 2008), and industrial societies (Emmott et al., 2020; Ishino, 1953) often find that non-kin, such as friends, neighbors, and fictive kin, provide allomaternal care. In industrialized, urban societies, paid caretakers, nannies, or daycare or school teachers often serve as allomothers, trading their skills for wages (Mayall, 2009).

As is true of allomaternal care by kin, non-kin allomaternal investment is diverse and can be broken into two general categories: direct care, which goes to the child, and indirect care, which goes to the mother. Direct care often involves holding infants or supervising children—for example, when older children supervise younger children in mixed-age playgroups (Page et al., 2021). Among Hadza, non-kin hold infants 12.4% of the time, more than any other single category of allomaternal kin caregivers (Crittenden & Marlowe, 2008), while among Agta, toddlers were within three meters of non-kin as frequently as they were siblings (Page, 2016). Indirect care may involve supporting or provisioning the child's parent, e.g., by supporting the mother as she breastfeeds (Emmott et al., 2020) or helping produce food (Jaeggi, Hooper, Beheim, Kaplan, & Gurven, 2016; Jang et al., 2022).

Whether non-kin allomaternal investment provides direct and indirect benefits is an open question, partly because most evidence is correlative and partly because there are many different outcomes to measure among mothers and children. That said, there is suggestive evidence that non-kin allomaternal investment may increase fertility for mothers, improve child health outcomes, and improve child educational outcomes. For example, mothers with more social support may have higher fertility (Shaver et al., 2020) and healthier children (Kana'iaupuni, Donato, Thompson-Colón, & Stainback, 2005), while social safety nets, like paid parental leave provided by employers or governmental bodies, can reduce infant mortality (Ruhm, 2000). The presence of kin can actually reduce support in some contexts, reducing financial support for mothers in Mexico (Kana'iaupuni et al., 2005) and reducing child survivorship in Mpimbwe, Tanzania (Borgerhoff Mulder & Beheim, 2011); non-kin may act as potential substitutes for kin allomothers in these contexts. Some evidence suggests that the presence of non-kin allomotherts is associated with children's educational outcomes, e.g., improved performance on cognitive tests (Shaver et al., 2020) and improved educational outcomes in young adults (Fruiht & Wray-Lake, 2013; Hagler & Rhodes, 2018; Hurd, Sánchez, Zimmerman, & Caldwell, 2012) and adult children (Perry, 2017). In short, though we have little direct evidence of the impact of non-kin investment on maternal and child outcomes, the evidence we do have suggests both direct and indirect benefits for children.

The question remains of why non-kin allomothers invest in mothers and children. Common explanations include reciprocity and fitness interdependence, coercion and other forms of manipulation, and learning to mother (see also Crittenden & Marlowe, 2008; Page et al., 2019). First, childcare is one of many currencies involved in reciprocal cooperation (Jaeggi, Hooper, Kaplan, & Gurven, 2014); non-kin allomothers may give direct or indirect care to children and receive delayed benefits in a different currency (Crittenden & Marlowe, 2008; Page et al., 2019). Further, if alloparents have fitness interdependence with the mother or child—that is, if their persistence may benefit the alloparent in the future—they may invest in one or both (Aktipis et al., 2018; Connor, 1986). Such fitness interdependence may be especially important when candidate kin cooperators are not available or not reliable cooperative partners (David-Barrett & Dunbar, 2017; Santos-Granero, 2007), or when the mother is in need (Snopkowski & Sear, 2016). Second, if individuals can penalize others for not helping (e.g., by withholding cooperation or inflicting reputational damage), allomothers may have little choice but to help (Crittenden & Marlowe, 2008). Relatedly, as described above, individuals may manipulate kinship terminology to elicit kin-selected behavior from alloparents (see Qirko, 2013 for discussion)—although data suggest that while fictive kinship promotes cooperation, it does not raise levels of cooperative behavior to that observed among kin (David-Barrett & Dunbar, 2017; Hruschka, 2010). Third, young allomothers may benefit from practicing mothering with unrelated children (Crittenden & Marlowe, 2008; Lancaster, 1971; Page et al., 2019).

In theory, then, strengthening non-kin relationships can enhance allomaternal investment in children and mothers, a strategy that may be important when candidate kin alloparents are unavailable (e.g., Kana'iaupuni et al., 2005; Shaver, 2017) or, given locally relevant risks to food security, when kin alloparents alone cannot provide adequate help (Page, 2016; Page et al., 2019). A principal means of strengthening these relationships is thought to be through the use of kinship terms (Ebaugh & Curry, 2000). Examples include the use of terms such as “brother” and “sister” in the Franciscan Family (Büssing, Recchia, & Dienberg, 2018), the application of “Aunt(y)” to women of similar age or socio-economic status as one's mother in many North Indian communities (Freed, 1963), or the use of “Aunt” or “Uncle” when addressing any adult in many African American communities (Chatters, Taylor, & Jayakody, 1994). We and other authors refer to this as fictive kinship, where individuals are not related to the mother or child by blood or marriage but are given familiar names and treated as an extension of the family (Ebaugh & Curry, 2000; Freed, 1963).

A category of fictive kin are ritual kin—non-kin (or sometimes kin) with specific rights or obligations established through a ceremony or ritual (Ebaugh & Curry, 2000; Freed, 1963)—and ritual kin can be particularly important alloparents. Because of the benefits that ritual kin can provide, many institutions have norms and/or rules for such ceremonies, in which an individual publicly commits to a cooperative relationship with the parent and/or child (see Ebaugh & Curry, 2000 for a review). By publicly committing to such a relationship, ritual kin can be sanctioned for later reneging on their duties, both of which support continued cooperation between ritual kin and children or mothers (Ebaugh & Curry, 2000; Foster, 1969). Indeed, relationships between ritual kin and children or mothers can last well beyond childhood (Freed, 1963). Touchstone examples of ritual kinship include “milk kin,” traditionally recognized by Islamic law (Altorki, 1980) and found in traditional Christian populations like the Romans, Jacobite Syrians, Armenians, and Copts (Parkes, 2005); the oyabun-kobun (“father-child”) system practiced in recent centuries in Japan (Ishino, 1953); and the practice of naming godparents in many Christian countries, called compadrazgo in Latin America and other former colonies of Spain (Gill-Hopple & Brage-Hudson, 2012; Mintz & Wolf, 1950).

In this paper, we focus on compadrazgo in a Latin American context. Compadrazgo creates networks of kin and non-kin with reciprocal obligations for resource access and cooperative support (Foster, 1969). Godparents (padrinos) are often considered second parents, responsible for the well-being of their godchild (ahijado) and the child's parents (their compadres; Foster, 1969; López, 1999; Mintz & Wolf, 1950). When they live in the same community as their godchildren and compadres, godparents may provide direct care by watching godchildren and provisioning food, indirect care by providing their compadres with resources or services (Foster, 1969; Gill-Hopple & Brage-Hudson, 2012), or loans to either their godchildren or their compadres (Chatters et al., 1994; Ebaugh & Curry, 2000; Gill-Hopple & Brage-Hudson, 2012). When godparents live in a different community, they may provide direct or indirect care to godchildren or compadres by providing them with access to non-local resources—like medical facilities, which can reduce mortality rates (Gill-Hopple & Brage-Hudson, 2012). If godchildren or compadres move to the godparent's location, the godparent may provide housing and help them establish social ties and locate resources (Ebaugh & Curry, 2000; O'Connor, 1990; Vidal, 1988). Given these potential benefits, parents often choose godparents for their children who they believe can best provide social support or resource access—whether to their children or to the parents themselves; for this reason, godparents are often from a different location or higher social class than their godchildren (Foster, 1969; López, 1999; Mintz & Wolf, 1950; van den Berghe & van den Berghe, 1966).

Here, we highlight field data from two Indigenous communities in rural Bolivia that practice compadrazgo, one Mosetén and the other an Intercultural (multicultural) community; we refer to the communities as “Mosetén” and “Intercultural” as pseudonyms. Both have populations of approximately 1000 individuals and engage in slash-and-burn horticulture, growing cash crops purchased by middlemen that are then sold in the capital city, La Paz, or in local markets. Both have access to roads (including transportation to local markets), running water, cell service, and electricity. Both practice adoption and fosterage, especially by kin. Both intermarry with members of different ethnic groups and neither practices polygamy. As of 2012, average completed family size was seven for Mosetén and five for Intercultural, although access to free birth control means that families are decreasing in size.

Both communities are largely Catholic and compadrazgo is predominantly a Catholic practice. Godparents are named at many different ceremonies, including baptism, rutucha (the first cutting of hair), kindergarten graduation, confirmation, quinceañera, and high school graduation. The majority are non-kin, and godparents can live inside or outside of the community. In the 2010s, Evangelical churches (e.g., Assembly of God and Church of the Nazarene) became important in Mosetén and especially Intercultural. Evangelical congregations in these communities discourage compadrazgo, instead relying on kinships terms like “brother” and “sister” with fellow members of their congregations. However, because many are recent converts, they often have compadres for their children from before their conversion. Fifty-two percent of children whose parents identify as Evangelical have at least one godparent.

Godparents are often important alloparents in Intercultural and Mosetén. Godparents in the community sometimes feed, babysit, and buy gifts for their godchildren, in addition to loaning their compadres money. Godparents outside of the community also may provide loans, but often also act as brokers for godchildren when they come to visit, providing them with a place to stay and helping them navigate local challenges, like bureaucracy and registering for school.

Primary and secondary schools are available in Intercultural and Mosetén. Intercultural has had a kindergarten through 12th-grade school (primary and secondary school; K-12 for short) since the early 1970s. Mosetén had a K-12 school nearby for a few decades, which most children attended, but a new K-12 school was opened in the community in 2010. While families do not pay tuition, there are costs of attending K-12 school, including uniforms, school supplies, and communal work like clearing the soccer field and preparing food for special events. It is expected that most children will finish high school, however children still drop out if they get pregnant or wish to work.

Children increasingly pursue various higher-education opportunities. Those with certificates or degrees in relevant fields can expect to make more money than they can in cash cropping; for example, those with agroforestry degrees often end up doing stints with oil companies. When they pursue higher education, children must move away from the two communities, usually either to La Paz (e.g., for a bachelor's degree or to study mechanics), which is six hours away by car, or nearby university extensions (e.g., for nursing or agroforestry), which are one hour away. University tuition costs B500 (approximately 73 USD) per term, equivalent to about seven days' wages in Mosetén and Intercultural; however, the largest expenses for students are living expenses. Students often stay with family members, close friends, or godparents to save money (Pisor & Jones, 2021; Pisor & Ross, 2022).

As aforementioned, fictive kin can play an important role in children's educational achievement and attainment (Hurd et al., 2012; Perry, 2017; Shaver et al., 2020). In rural Bolivia, per our review above, this may especially be the case given godparents provide both direct and indirect care in communities—like childcare and loans that can pay for K-12 school costs—and out of communities—like a place to stay and help with bureaucracy for students pursuing higher education. We investigate whether godparents have notable impacts on children's outcomes in this context by evaluating the following predictions, each for children over 18 and each controlling for the presence of kin alloparents (see Fig. 1):

P1: Adult children with one or more godparents will be more likely to have (a) completed more years of primary and secondary education, (b) finished high school, and (c) pursued higher education than children without godparents—that is, they will have higher educational attainment.

P2: As a check to ensure this effect is not an effect of kin disguised as an effect of godparents, we predict that both kin godparents and non-kin godparents will have impacts on the educational attainment (a, b, c above) of adult children.

P3: Godparents in the community will predict increased (a) completed years of primary and secondary education, and godparents outside the community will predict (c) pursuit of higher education.

Further, given their known impacts on child outcomes, we explore whether the presence of kin has direct effects on educational attainment (Fig. 1).