Serbian Roma: Girls were the preferred sex, owing to expected fitness benefits; Roma fathers tend to bias their investment towards taller, more endowed children, because of greater fitness pay-off

Paternal investment, stepfather presence and early child development and growth among Serbian Roma. Jelena Čvorović. Evolutionary Human Sciences, Vol 4, April 18 2022. https://www.cambridge.org/core/journals/evolutionary-human-sciences/article/paternal-investment-stepfather-presence-and-early-child-development-and-growth-among-serbian-roma/F3B5185E6333B85FEF0ACCFD5685FCA4

Abstract: Research on paternal investment and child growth and development is limited outside of high-income countries. Using nationally representative data from low-resource Serbian Roma communities, this study examined father investment (direct care), its predictors and the associations between paternal investment, stepfather presence and child physical growth and early development. The sample included 1222 children aged 35–59 months, out of which 235 were living with biological fathers. Child outcomes included height-for-age Z-scores, stunting and early child developmental score. Roma paternal investment was relatively low. There was a positive association of father investment and children's height, and no association with developmental score. The presence of father vs. stepfather did not exert any influence on children. Instead, maternal and child characteristics explained both the overall development and height for Roma children. Thus, older children, born to literate, lower parity mothers of higher status and greater investment had better developmental and growth outcomes; girls were the preferred sex, owing to expected fitness benefits. Reverse causality emerged as the most likely pathway through which the cross-sectional association of father direct care with child growth may manifest, such that Roma fathers tend to bias their investment towards taller, more endowed children, because of greater fitness pay-off.

Discussion

This paper used nationally representative data from Serbian Roma communities to assess paternal investment (as defined in this paper) and its predictors, and the associations between paternal investment, stepfather presence and early child development and physical growth.

There were several main findings. In this Roma sample, parental investment in a number of direct care behaviours was relatively low, while maternal and paternal investments were positively correlated. Overall, almost 10% of children did not receive any investment from their parents, as measured in the present study. In line with other studies from low- and middle-income countries, almost one-third of Roma fathers did not interact with their children in the surveyed care behaviours, twice the proportion of unengaged Roma mothers: 29.7 vs. 14.2% (Jeong et al., Reference Jeong, McCoy, Yousafzai, Salhi and Fink2016; Sun et al., Reference Sun, Liu, Chen, Rao and Liu2016). There was a cross-sectional association between father's investment with child's height and stunting, while maternal investment and child's height appeared to be predictors of father's investment. There were no associations between father's investment and developmental score, and stepfather vs. father presence and Roma children outcomes. Instead, maternal and child characteristics explained both the overall development and height for Roma children.

Parenting practices may be influenced by a range of child and family characteristics, and political and economic development (Walker et al., Reference Walker, Wachs, Grantham-McGregor, Black, Nelson, Huffman and Richter2011). In addition, local culture and traditions influence parenting behaviour: in many countries, childcare is culturally ‘mother centric’, with low participation from fathers (Hosegood & Madhavan, Reference Hosegood and Madhavan2010). Thus, for instance, an average Serbian father spends only 11 minutes per day with his children, while only one in 20 fathers is fully involved in parenting (Republički zavod za Statistiku, 2020). The observed low paternal investment of Roma fathers may reflect the prevailing dominant patriarchal cultural pattern with a significant sex asymmetry in parenting (Čvorović & Coe, Reference Čvorović and Coe2019). Additionally, bias in maternal reporting could account for this finding – studies have found that generally fathers tend to report significantly higher levels of involvement than mothers, contingent on numerous factors including ethnicity, the quality of the couple's relationship and the child characteristics (Charles et al., Reference Charles, Spielfogel, Gorman-Smith, Schoeny, Henry and Tolan2018).

Previous research found inconsistent relationships between father's direct care and children's height (Maselko et al., Reference Maselko, Hagaman, Bates, Bhalotra, Biroli, Gallis and Rahman2019; Jeong et al., Reference Jeong, McCoy, Yousafzai, Salhi and Fink2016), but in this study, after adjusting for potential confounding factors, Roma fathers’ direct care was positively associated with their children's height. Fathers may contribute to their offspring's well-being in a number of ways: additive paternal care (‘cooperative care’ where both parents work together to care for children at the same time) can include complementing the mother's direct care or providing resources that allow for better nutrition (Gurven et al., Reference Gurven, Winking, Kaplan, von Rueden and McAllister2009). Additive care may also include playing with a child or teaching (Starkweather et al., Reference Starkweather, Keith, Prall, Alam, Zohora and Emery Thompson2021). Paternal additive direct care may have similar impact to that of other allomothers: a child may receive a better overall gain and additive investments that thus could lead to better fitness outcomes (Emmott & Page, Reference Emmott, Page, Shackelford, A and Weekes-Shackelford2019). Given the positive correlation of Roma maternal and paternal investment, the overall investment may have positively affected children, resulting in better outcomes. However, the most obvious mode that paternal direct care can influence growth is via nutrition, i.e. feeding. MICS is cross-sectional in design, and measures of direct parental involvement did not include feeding practices, thus the results may be confounded by an unmeasured variable that correlates with paternal care. In this setting, another possible pathway that may reflect on the positive association between father investment and child height is reverse causality, such that fathers provided more care to taller/healthier children (Maselko et al., Reference Maselko, Hagaman, Bates, Bhalotra, Biroli, Gallis and Rahman2019). A Roma child's height, in addition to maternal investment, predicted father's investment: taller children had fathers who provided more direct care. In young children, height serves as a proxy for the cumulative effect of nutritional and health loads from conception (Frongillio et al., Reference Frongillo, Kulkarni, Basnet and de Castro2017). Thus, early childhood growth is an important measure of offspring quality, as it may influence future health and reproduction (Kramer et al., Reference Kramer, Veile and Otárola-Castillo2016).

Additionally, Roma child's height and the chances of stunting were influenced by the child's sex: Roma boys were more likely to be shorter and stunted than Roma girls, suggesting that they were more susceptible to nutritional inequalities than their girl counterparts of the same age. This pattern is consistent with previous findings where biased paternal investment was associated with the children's health/height but also sex (Alvergne et al., Reference Alvergne, Faurie and Raymond2009; Čvorović, Reference Čvorović2020a; Hagen et al., Reference Hagen, Hames, Craig, Lauer and Price2001). Among the Roma, parents may selectively invest in and support taller children (girls) who had the greatest potential to survive into adulthood and reproduce successfully, thus making the parents into grandparents, or in other words, enhancing the parents’ reproductive success (Čvorović, Reference Čvorović2020a). Sex preferencing among the Roma in favouring girls is a common finding in Hungarian Roma groups as well (Bereczkei & Dunbar, Reference Bereczkei and Dunbar2002). Roma girls more often than boys engage in helping-at-the-nest, have a greater chance of marrying up the socio-economic scale and produce more surviving children compared with sons. In addition, having a high-quality (taller/healthier) daughter is regarded as a considerable advantage and a source of income among Serbian Roma who practice bride price (Čvorović, Reference Čvorović2014).

Compared with other studies, where socioeconomic status and parental education were positively associated with early child development (Urke et al., Reference Urke, Contreras and Matanda2018; Jeong et al., Reference Jeong, McCoy and Fink2017; Paxson & Schady, Reference Paxson and Schady2010), in this setting, there was an apparent lack of relationship between child developmental score and also four domains of development (literacy–numeracy, physical, socioemotional and learning) with paternal investment and other variables.

In contrast to other studies, when a stepfather was introduced to the context, the presence of either father or stepfather had no influence on Roma children outcomes (Case & Paxson, Reference Case and Paxson2001). Instead, child's age and maternal characteristics explained both the height and overall development for Roma children. Thus, older children were taller and had higher developmental score. Older children have higher reproductive value, and in poor populations, the later born children are often disadvantaged relative to earlier borns in nutritional status and growth, having higher morbidity and mortality (Lawson & Mace, Reference Lawson and Mace2009). In addition, maternal parity was negatively associated with child's height. Roma mothers with higher parity had children who were shorter than those of mothers with lower parity. Under poor conditions, numerous siblings may put children at higher risk of malnutrition, because of the discrepancy between family size and available resources. Maternal parity may also serve as a rough measure for investment: body size may be a proxy for a trade-off between offspring number and quality, or between the number and size of offspring, especially under resource-scarce conditions and in high-fertility settings such as with the Roma (Walker et al., Reference Walker, Gurven, Burger and Hamilton2008). Maternal parity was higher in stepfather households, but interestingly, unlike in other studies (Amato & Rivera, Reference Amato and Rivera1999; Lawson & Mace, Reference Lawson and Mace2009), maternal investment was higher for children living with stepfathers. One possible explanation as of why children living with stepfathers experience higher maternal investment may be that a stepfather is providing some extra resources, thus the mother may be experiencing a higher quality of life in this new relationship, both of which enable the mother to better provide for her children, including direct care. MICS does not include information on stepfathers’ resources, while maternal age, basic literacy skills and access to improved sanitation (as proxies for socioeconomic status) did not differ between mothers living with biological fathers vs. stepfathers. A more likely explanation is that Roma mothers in stepfather households may be compensating for the absence of a biological father by focusing more investment and attention on the children from former unions (Emmot & Mace, Reference Emmott and Mace2014). Higher investment and more attention could also serve as protection for any number of possible negative effects in the new home. This context could perhaps explain the lack of association of maternal and paternal characteristics and child development and maternal characteristics and child height: the poor conditions may have affected the ability of Roma parents to invest, and to make substantially enough investment to be detected or differentiate between children. Thus, mothers ‘get activated’ only in the presence of stepfathers (high risk), to protect their children and compensate for even the limited paternal investment. Roma children may be sensitive to this particular setting as well: child's age was associated with growth and development only in the stepfather's presence, thus younger and older sibling get to compete more in a stepfather's household, as there is an actual maternal investment to compete for.

Furthermore, maternal age was positively associated with both height and development: generally, older mothers tend to invest more in offspring, as they are less likely to have additional children and the investment focuses on the children they already have (Uggla & Mace, Reference Uggla and Mace2016). An additional explanation could be that this relationship reflects on maternal status within a family. Many Roma women face inflexible gender roles and expectations, and for many, having children in marriage is the only socially endorsed route for an improvement in status (Čvorović & Coe, Reference Čvorović and Coe2019). A higher maternal status within a Roma family may include more power in decision-making concerning child's wellbeing such as diet and activities.

Additionally, there was a positive association of child's age, maternal investment, and literacy with children's overall developmental score. As a child ages, it is more likely to develop and learn skills and be ahead in development. The importance of maternal care behaviours and education for children's early development has been well described: parental support for learning (such as stimulating interactions and reading books) was found to be an important means through which parental education is associated with children's development (Sun et al., Reference Sun, Liu, Chen, Rao and Liu2016; Jeong et al., Reference Jeong, McCoy and Fink2017; Walker et al., Reference Walker, Wachs, Grantham-McGregor, Black, Nelson, Huffman and Richter2011). In turn, maternal education can facilitate maternal investment and practices, as increasing levels of education lead to different thinking and decision-making patterns (Cutler & Lleras-Muney, Reference Cutler and Lleras-Muney2010). This may be especially important to the Roma, given the high illiteracy rate among females: even low levels of education increase children's well-being and survival prospects (Sandiford et al., Reference Sandiford, Cassel, Sanchez and Coldham1997; Čvorović, Reference Čvorović2020a).

To the best of my knowledge, this is the first study to provide an account of paternal direct care as a proxy for investment, stepfather presence and child development and growth among the low-resource Roma. The study findings contribute new evidence of the drivers or lack of it of development and growth among children in marginalised ethnic populations, adding to the literature about paternal investment and child outcomes.

The majority of Roma children grow up in poverty, born to mothers with low education, and in homes with limited learning opportunities. In this context, parental investment was relatively low. Fathers have limited involvement in direct care of their young children and this involvement was not associated with child development. The presence of a father vs. a stepfather did not exert any influence on Roma children, insomuch as it did not have direct influence on the children's’ outcomes of growth and development. Roma paternal investment was low to begin with and father absence is likely to be less important in settings where fathers usually provide less support for their children (Lawson et al., Reference Lawson, Schaffnit, Hassan, Ngadaya, Ngowi, Mfinanga and Borgerhoff Mulder2017). In the presence of a stepfather, maternal and child's traits explained overall child development and growth. Maternal investment was higher for children living with stepfathers, thus mothers may be protecting their children from previous unions and compensating for paternal absence. Competition among Roma children – among younger and older siblings – was evident only when maternal investment was significant, in the presence of a stepfather. Thus, older children, born to literate, lower parity mothers of higher status and greater investment had better developmental and growth outcomes; girls were the preferred sex, as they were likely to be taller and less stunted than Roma boys, possibly owing to expected fitness benefits. Reverse causality emerged as the most likely pathway through which the cross-sectional, positive association of father direct care with child growth may manifest, such that Roma fathers tend to bias their investment towards taller, more endowed children, because of greater fitness pay-off.

The study had several limitations. The data were cross-sectional, limiting causal inferences between the variables under study. The developmental score and paternal investment were mother-reported and thus subject to biases: both measure how mothers perceived their child's development and their husbands’ involvement, and not actual child development and paternal behaviour. Furthermore, the reliability of Early Child Development scale was fair but similar to other recent studies, reflecting its limited adaptation to local culture and context (Urke et al., Reference Urke, Contreras and Matanda2018; McCoy et al., Reference McCoy, Sudfeld, Bellinger, Muhihi, Ashery, Weary and Fink2017). The questions regarding literacy/numeracy have been shown to be too advanced for 3- and 4-year-old children (McCoy et al., Reference McCoy, Peet, Ezzati, Danaei, Black, Sudfeld and Fink2016), this being particularly pertinent as regards Roma and other disadvantaged children where parental literacy is limited.

Additionally, to date, no specific growth references have been developed for the Roma, even though their Indian origin indicate an ethnicity impact to the anthropometric measures. Albeit the population-specific growth references may serve as a more biologically relevant measure of within-population assessment of children's growth (Kramer et al., Reference Kramer, Veile and Otárola-Castillo2016; Martin et al., Reference Martin, Blackwell, Kaplan and Gurven2019), the effects found in this study may be considerable, ranging from an approximately 0.33 SD in child's height to a more than 1 SD difference in paternal care (Winking & Koster Reference Winking and Koster2015).

As child's growth and development are sensitive to available resources, and may be affected by aspects outside of direct family influence (Lawson et al., Reference Lawson, Schaffnit, Hassan, Ngadaya, Ngowi, Mfinanga and Borgerhoff Mulder2017; Winking & Koster, Reference Winking and Koster2015), social assistance (cash transfers) may also have an effect on Roma family, including the growth and development of Roma children. For instance, in affluent settings, fathers tend to engage more in direct child care if their wives are employed and/or contribute a greater share of the couple's earnings (Raley et al., Reference Raley, Bianchi and Wang2012). Roma mothers’ receipt of welfare could motivate Roma fathers to engage in direct child care: the majority of Roma women do not work (formal income leads to withdrawal of social benefits), but still support the family with cash transfers. Nevertheless, a recent study found that among Serbian Roma, receiving social assistance was associated with disintegration and a diminished role of the family (Čvorović & Vojinović, Reference Čvorović and Vojinović2020), but whether welfare influence father–child relationships and child outcomes remains unexplored. Also, other potential confounders, such as parental height and health status, were not collected. To fully understand the effects of paternal investment on child outcomes, information should be collected directly from fathers and father-like figures and/or through observation, and include parental anthropometrics, as well as data on the presence of alloparents, which may have an effect on child outcomes, including growth (Sear & Mace, Reference Sear and Mace2008).