The Role of Fathers in Promoting Early Childhood Development in Low- and Middle-Income Countries: A Review of Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 the Evidence David K. Evans and Pamela Jakiela The role of fathers in parenting young children differs dramatically across societies. Policy- makers in low- and middle-income countries (LMICs) are increasingly focused on increas- ing the amount of early childhood stimulation and other investments received by young children in an attempt to increase human capital accumulation, promote school readi- ness, and improve long-run outcomes. Until recently, few interventions targeted fathers, and most impact evaluations implicitly assumed that fathers played no meaningful role in parenting. We survey the emerging literature on fathers’ involvement in early childhood stimulation and parenting in LMICs and summarize the patterns of results from rigorous impact evaluations. We find that fathers spend less time stimulating their children’s devel- opment than do mothers or other adults who live in the same households in almost every country in our sample, and that mothers’ and fathers’ investments are positively associ- ated. We also find evidence that fathers’ take-up of parent training programs tends to be low, but that such programs have had effects on changing fathers’ knowledge and—to a lesser degree—behaviors. JEL Codes: J13, J16, J24, O15 Keywords: early childhood development; parenting; fathers; gender roles, household income. Introduction Hundreds of millions of young children in low- and middle-income countries (LMICs) are at risk because of a lack of adequate nutrition and early childhood stimulation The World Bank Research Observer © The Author(s) 2024. Published by Oxford University Press on behalf of the International Bank for Reconstruction and Devel- opment / THE WORLD BANK. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. https://doi.org/10.1093/wbro/lkae009 40:211–228 (Black et al. 2017). Well-designed early childhood development (ECD) interventions can have substantial impacts on children’s physical, cognitive, and socioemotional de- velopment, as well as their eventual schooling attainment, wages, and other life out- comes (Engle et al. 2011, Gertler et al. 2014, Black et al. 2017). As evidence of the ef- fectiveness of ECD interventions mounts, both governments and international institu- tions are investing in preprimary education and other early childhood development programs.1 Early childhood interventions often have substantial impacts on family Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 members other than young children, and a number of common policy interventions— for example, parenting education classes—only impact children by changing the behav- ior of adult household members (Evans, Jakiela, and Knauer 2021). Though the impacts of the most common early childhood interventions on young children and their pri- mary caregivers are reasonably well-documented (Evans, Jakiela, and Mendez Acosta 2024, Halim, Perova, and Reynolds 2022), there is comparatively little evidence on how such interventions impact other household members, or the roles that these individu- als play in the lives of children. We review the evidence on the role of fathers in child development during the first six years of life, focusing primarily on evidence from LMICs. Though rarely explicitly defined, ECD typically refers to the wholistic process of physical, cognitive, emotional, and social development in early life. This stage of life includes critical and sensitive pe- riods, when positive and negative shocks (e.g., changes in nutrition or cognitive stimu- lation) may be more likely to have persistent impacts on later-life outcomes (Black et al. 2017). Development in early childhood is supported through the provision of “nur- turing care,” which can be defined as “a stable environment that is sensitive to chil- dren’s health and nutritional needs, with protection from threats, opportunities for early learning, and interactions that are responsive, emotionally supportive, and de- velopmentally stimulating” (Britto et al. 2017). Nurturing care encompasses caregiving practices including feeding and hygiene, cognitive stimulation, responsive parenting, and safety. For most young children, mothers are the main agents providing nurturing care, but fathers, other relatives, childcare professionals, teachers, and other commu- nity members may also play a role. Fathers’ role in parenting is often ignored in discussions of early childhood care and policy. There are many reasons for this. Most obviously, fathers do less active childrea- ring than mothers in most societies, and they are particularly uninvolved in the care of young children (Guryan, Hurst, and Kearney 2008, Lancy 2015). From a theoretical perspective, many models of household specialization suggest that this pattern reflects an efficient allocation of household resources: men specialize in productive work away from home, or in settings that might be unsafe for children, while women specialize in home production, pregnancy, breastfeeding, and childcare (Boserup 1970, Becker 1981). Yet, this pattern is changing over time as total fertility declines and economies shift from “brawn-based” models of production to “brain-based” models of production (Pitt, Rosenzweig, and Hassan 2012). In rich countries, fathers do more than they did 212 The World Bank Research Observer, vol. 40, no. 2 (2025) a generation ago (Doepke and Zilibotti 2017). However, their role in caring for young children in LMIC contexts is still little scrutinized, and many early childhood interven- tions implicitly assume that parenting young children is the sole responsibility of the mother (e.g., by targeting mothers for parent training interventions). We review the evidence on fathers’ roles in early childhood development in LMICs in two steps. First, we use data from UNICEF’s Multiple Indicator Cluster Surveys (MICS) to characterize cross-country variation in fathers’ involvement in early childhood stim- Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 ulation. This updates and expands earlier analysis of paternal engagement using MICS data (Jeong et al. 2016; Cuartas et al. 2020). Stimulation is only one dimension of nur- turing care. However, in the absence of internationally comparable data on time use by mothers and fathers in LMICs, the MICS measure of stimulating activities by each parent provides a unique window into the division of parenting responsibilities within the household. We find that fathers consistently spend less time stimulating their chil- dren’s development than do mothers or other adults who live in the same households. We also find that mothers’ and fathers’ investments tend to be positively associated: countries with high rates of early childhood stimulation by mothers are also countries with high rates of early childhood stimulation by fathers. We then draw upon a systematic review of impact evaluations of ECD interventions in LMICs to summarize the state of the literature on fathers’ parenting and early child- hood development. We identify ten parenting interventions that explicitly target fa- thers as participants (e.g., in parenting classes), either on their own or together with their spouse. Studies of parent training that include fathers do show changes in fathers’ knowledge and some changes in behaviors but not in others. Adding fathers to inter- ventions that are primarily designed for mothers has more limited impacts: although there are examples of positive impacts, take-up by fathers is often low. We also iden- tify five evaluations of parenting interventions that did not explicitly target fathers, but where impacts on fathers’ parenting practices were measured at endline. These stud- ies suggest that in most cases, parenting interventions targeting the mother or primary caregiver do not generate detectable spillovers onto fathers’ parenting behaviors. Ultimately, we document clear evidence that fathers invest significantly less in child stimulation activities than mothers, and we find limited but promising evidence that there are ways to boost paternal engagement in child stimulation through parent train- ing. However, the total effect on a wider array of aspects of fathers’ lives and the impact of other efforts to engage fathers remains an area requiring much further study. Like- wise, because most of these interventions target fathers along with other family mem- bers, it is difficult to isolate the impact of boosting fathers’ involvement on children’s outcomes. A previous effort to measure the overall effect of father-inclusive interven- tions (not just the impact of directly targeting fathers for engagement) on child devel- opment found very heterogeneous results (Jeong et al. 2023a). Our systematic review also allows us to document the extent to which studies of other ECD interventions—for example, center-based childcare and targeted cash Evans and Jakiela 213 grants—impact fathers. There are strong theoretical reasons to expect that many ECD interventions will have direct impacts on household budget constraints and shift the distribution of domestic responsibilities. When this is the case, fathers may be affected—even if they spend little time on childcare. To date, relatively few studies re- port the impacts of ECD interventions on fathers, but many of those that do find sub- stantial impacts of policies such as center-based care on fathers’ labor supply and in- come. Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 The Role of Fathers in Parenting Young Children Fathers’ role in parenting differs dramatically across societies. In every society, moth- ers do more active childcare than fathers, but fathers’ involvement ranges from near zero to almost equal to that of mothers (Lancy 2015). The causes and consequences of fathers’ involvement in parenting are not well understood. One mechanical reason that fathers are less involved in childcare than mothers is that they are less likely to be present. UNICEF’s Multiple Indicator Cluster Surveys (MICS) capture a broad range of human development outcomes related to women and children, and provide a unique source of internationally-comparable data on parent- ing practices across a broad range of LMICs. Like the Demographic and Health Sur- veys, the MICS collect data on representative samples of young children; however, the MICS are unique because, for a subset of countries, they also contain information about parenting practices and early childhood stimulation (as discussed further be- low).2 Across 69 countries for which data on parenting is available, the average share of children who live with their mother is over 90 percent, while the average share of children who live with their father is only 72 percent. The share of children living with their father also varies considerably across regions, from 61 percent in Latin America and the Caribbean to over 96 percent in the Middle East and North Africa.3 The reasons that fathers are absent also vary: in some cases, male household members migrate and send remittances home, so an absent father may translate directly into better welfare outcomes for the household (Mansuri 2006; Munshi 2020). Consistent with this, the empirical evidence on the developmental consequences of fathers’ absence is, at best, mixed. In a review of the impacts of fathers’ presence in the home on child survival, Sear and Mace (2008) note, “Fathers have surprisingly little effect on child survival, with only a third of studies showing any beneficial effects.” Their conclusion reflects two main strands of literature on father absence, one exploring the impact of having a father die (i.e., becoming a paternal orphan) and a second exploring the impact of father migra- tion (i.e., when the father is still alive but no longer lives in the same community). This literature includes outcomes for children beyond early childhood. Literature on paternal orphans in LMICs has tended to find that losing one’s father does not have dramatic effects on child outcomes. Studies in Kenya and South Africa found no impact of father deaths on children’s school participation 214 The World Bank Research Observer, vol. 40, no. 2 (2025) (Case and Ardington 2006; Evans and Miguel 2007). In Tanzania, there is a modest ef- fect of paternal death on schooling if the father lived in the household at the time of death, but the impact is much smaller than the effect of maternal death (Beegle et al. 2010). The migration literature also finds mixed results. In Mexico, paternal migration is as- sociated with worse child health outcomes, both for any illness and for diarrhea specif- ically (Schmeer 2009). Likewise in Guatemala, international migration of a father in the Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 previous year is associated with lower height-for-age scores for children, even control- ling for remittance income, perhaps because fathers who migrated one year ago have not yet begun remitting income to their families (Davis and Brazil 2016). In contrast, a study in Mozambique identifies no difference in mortality rates across migrant and non-migrant men’s children (Yabiku, Agadjanian, and Cau 2012). For education out- comes, instrumental variables estimates in China suggest a modest adverse effect on children’s test scores when fathers migrate, though there is a much larger negative ef- fect when mothers migrate (Zhao et al. 2014), while Li, Liu, and Zang (2015) do not find that migration by either parent impacts child health. Taken together, the existing body of evidence suggests that negative impacts of fathers’ absence on young children tend to be modest in most contexts. Even when they are present, fathers are typically less involved in the daily ac- tivities of parenting than other household members. Figure 1 presents MICS data on early childhood stimulation by mothers, fathers, and other household members. The surveys capture the extent of early childhood stimulation that children receive at home through the Family Care Indicators Questionnaire (Hamadani et al. 2010, Kariger et al. 2012), which asks about six different types of stimulating activities (e.g., shared reading, storytelling, and physical play). For each activity and each young child in the household, enumerators record whether an adult household member engaged in the activity with the child in the 72 hours prior to the survey. This provides a child-level measure of early childhood stimulation experienced, but also a breakdown of the extent to which mothers and fathers share responsibility for engaging with their young children. Across all 69 countries, mothers engage in an average of 2.9 (out of 6) stimulating activities with young children, while fathers engage in only 1.3 stimulating activities. Mothers engage in more stimulating activities than fathers in every country except Thailand, where mothers engage in an average of 1.1 stimulating activities with each child while fathers engage in 1.3. Across countries, fathers also engage in fewer stimu- lating activities with young children than other adult household members, who engage in an average of 1.6 activities with each child. Figure 1 plots the country-level averages and compares the extent of early childhood stimulation by fathers to what is done by mothers and other adult household members. The left panel of the figure demonstrates that the number of stimulating activities done by fathers and the number done by mothers are positively correlated, increasing more Evans and Jakiela 215 Figure 1. Early Childhood Stimulation by Fathers, Mothers, and Others Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 Source: Authors’ analysis of data from Multiple Indicator Cluster Surveys in 69 Countries Note: Data from Multiple Indicator Cluster Surveys, rounds 4, 5, and 6 (collected between 2010 and 2023). Outcomes are based on responses to the Family Care Indicators (FCI) survey module, which asks about early childhood stimulation experienced by children aged 3 and 4 in the 72 hours prior to the survey. The FCI captures six distinct stimulating activities (singing songs, telling stories, playing, excursions outside the home, shared reading, and learning activities) and for each activity record the name of the adult household member who engaged in the activity with the child. or less linearly (correlation = 0.88). This pattern is not consistent with a model where either the mother or the father specializes in parental stimulation, while the other par- ent engages in other activities that also benefit the household (such as paid work out- side the home). The three countries with the highest reported rates of early childhood stimulation by mothers—Montenegro, Serbia, and Bosnia and Herzegovina—are three of the four countries with the highest rates of stimulation by fathers; and the three countries with the lowest rates of stimulation by fathers—The Gambia, Uzbekistan, and Lesotho—are among the four countries with the lowest rates of stimulation by mothers. The right panel of fig. 1 shows the relationship between stimulation by fathers and stimulation by other adult household members. The two appear to be largely unrelated (correlation = 0.18). One clear pattern that stands out in both figures is that rates of early childhood stimulation by both mothers and fathers, as measured in the MICS sur- veys, are substantially lower in Sub-Saharan African than in all other regions. The low levels of stimulation by both parents might be seen as evidence that African parents see stimulating activities as less important, developmentally. Such an interpretation res- onates with work (cf. Weber, Fernald, and Diop 2017, Jukes et al. 2018) suggesting that prevalent cultural norms may discourage African parents from conversing with their young children. However, the evidence on stimulating activities by other household members runs counter to this interpretation: fig. 1 illustrates that, while rates of early 216 The World Bank Research Observer, vol. 40, no. 2 (2025) childhood stimulation by African parents tend to be low, rates of cognitive stimulation by other adults in Sub-Saharan Africa are similar to those observed in other regions. This is consistent with anthropological evidence on the importance of alloparenting and child fostering in many African contexts (Akresh 2009, Lancy 2015). Importantly, neither panel of the figure suggests that early childhood stimulation by fathers substitutes for attention from other adult household members, since it is not negatively correlated with stimulation by either mothers or others. Instead, fathers’ Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 level of engagement in early childhood stimulation is highly correlated with mothers’ behavior, though mothers typically do between two and three times as much as fathers. The descriptive evidence from the MICS raises important questions about fathers’ role in parenting young children. One question is: why do fathers do so little early child- hood stimulation? Are they bad at it, in the sense that children gain less from respon- sive parenting by fathers than from responsive mothering or responsive alloparenting by other adults in the household? Alternatively, fathers may be constrained by cultural norms that assign childrearing responsibilities to women and girls (Lancy 2015). If this is the case, households may be misallocating fathers’ time toward more traditional or socially acceptable activities that have a lower return than engaging in responsive par- enting. A second question is whether parenting norms can be influenced by policymak- ers. Does providing parents, and specifically fathers, with information on the returns to early childhood investments change fathers’ parenting behavior? Can they be nudged to provide more nurturing care to their young children, or are norms related to division of labor within the household strong enough that encouraging fathers to do more will have little effect? A third question is whether convincing fathers to do more engaged parenting will increase child development and household welfare. Would increasing fathers’ involvement with their young children have positive impacts on early child- hood development, or is it better to target parenting interventions to women because they spend more time with young children? To explore these questions, and others, we surveyed the existing literature on early childhood interventions in LMICs that either attempted to change fathers’ parenting practices or measured impacts on fathers’ en- gagement with their children. Interventions Targeting Fathers’ Parenting Practices Our analysis builds on a systematic review of 3,716 ECD studies published between 2005 and 2019 (Evans, Jakiela, and Knauer 2021). That study identified 478 experi- mental and quasi-experimental evaluations of ECD interventions in LMICs, only 12 of which reported any results specific to the father or male head-of-household. We extend that sample by systematically identifying papers that cite the original studies, leading to the inclusion of ten more recent studies. Thus, we discuss a total of 22 impact evalu- ations of ECD interventions in LMICs that report father-specific results, either impacts on parenting practices or on other individual-specific outcomes such as earned income Evans and Jakiela 217 or mental health. The results of the 22 studies are summarized in Appendix Table A3.2; the search is characterized in more detail in Appendix A1. All 22 studies are of relatively high quality, with defensible identification strategies and levels of attrition. We discuss variation in study quality in Section 5 and Online Appendix A2. While it is still the case that most parenting interventions in LMICs only target mothers (or primary caregivers, who are typically female), the number of programs in- cluding or even targeting fathers has increased substantially in recent years. Our sam- Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 ple of 22 papers includes ten evaluations of interventions that either only target fathers or target fathers and mothers together. The first impact evaluation in our sample of an ECD intervention in an LMIC that explicitly targeted fathers was published in 2014, and only one impact evaluation study published before that even reported father-specific parenting results. However, our sample includes three evaluations that were published in peer-reviewed journals since 2020, plus two additional unpublished working papers (or preprints) that have been posted in the last three years, and several additional on- going randomized evaluations related to fathers’ parenting practices in LMICs are cur- rently listed in the American Economic Association trial registry. Hence, the body of evidence on fathers’ parenting practices—though small at present—in LMICs is likely to expand over the next few years. Because the overall number of father-focused ECD interventions is modest, it is still possible to provide a narrative overview of all existing studies. Doyle et al. (2018) evaluate a program targeting fathers with young children in Rwanda; study participants and their ( female) partners attended weekly group meet- ings encouraging reflection and dialogue around topics related to parenting and gen- der equality. The intervention reduced physical violence against both mothers and chil- dren, increased uptake of modern contraception and antenatal care, and increased women’s say in household decision-making, although it did not increase the amount of time men devoted to domestic work. Björkman Nyqvist and Jayachandran (2017) study a parenting intervention in ru- ral Uganda—classes on nutrition and child health—that was randomized at the village level to either mothers or fathers. They find that attendance rates were higher among targeted mothers (who attended 76 percent of sessions) than targeted fathers (who attended 58 percent of sessions). Intent-to-treat effects on the health knowledge of targeted parents were similar for the two interventions, but the intervention target- ing fathers led to health knowledge spillovers onto mothers (program participants’ wives), while the intervention targeting mothers did not improve their husbands’ health knowledge. However, only the arm that targeted women improved household- level health behaviors. Lighter touch parenting education for fathers has also proven effective in some con- texts. Dinga (2019) reports that providing the fathers of newborns with breastfeeding education in Kenya improves their breastfeeding knowledge, while Wang et al. (2015) find that counseling can help fathers of young children quit smoking in China. All four 218 The World Bank Research Observer, vol. 40, no. 2 (2025) studies demonstrate that ECD interventions explicitly targeted at fathers can be effec- tive at improving their parenting knowledge, and can also lead to changes in behavior that are likely to benefit children. However, the more ambitious studies by Björkman Nyqvist and Jayachandran (2017) and Doyle et al. (2018) also show that parenting norms can be sticky. Some parenting norms related to the division of labor within the household may be difficult to influence, even with relatively intense interventions; and in places where traditional norms persist, mothers may be better placed to act on par- Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 enting knowledge than fathers. Consistent with this, the evidence on adding fathers to parenting interventions pri- marily focused on mothers (or primary caregivers) is mixed. In a study in rural Kenya, Luoto et al. (2021) test whether encouraging fathers to participate in group-based par- enting education improves outcomes for children or mothers, but find that their inter- vention failed to convince fathers to regularly attend training sessions.4 Overall, tar- geting fathers did not impact outcomes for mothers or children in that context. Their results resonate with those of Justino et al. (2023), who found that encouraging Rwan- dan fathers to attend parenting classes with their wives led to a participation rate of just over 10 percent.5 In contrast, Özlüses and Çelebioglu (2014) find that explicitly targeting both fathers and mothers with breastfeeding education improves paternal attachment with infants in Northern Cyprus. Carneiro et al. (2021) find that providing information on child health to both mothers and fathers in northern Nigeria leads to increases in health knowledge for both parents, though impacts on mothers are larger and more persistent. Amaral et al. (2021) evaluate a digital parenting program in El Salvador that was im- plemented during the first year of the COVID pandemic. Both mothers and fathers were eligible to participate, and treatment assignments were stratified by (parent) gender. They find that the program increased male caregivers’ stress and anxiety and reduced the frequency of positive interactions with young children—though the intervention had no statistically significant impacts on female caregivers. There is also some evidence that parenting programs that do not explicitly target fathers can impact their childrearing practices. Jensen et al. (2021), Antelman et al. (2023), and Rahman et al. (2008) find that parenting interventions primarily targeted at mothers improved fathers’ engagement with their children in Tanzania, Rwanda, and Pakistan (respectively), though Bos et al. (2024) do not find effects of home vis- its from child development specialists on fathers’ parenting practices in Bangladesh. Osaki et al. (2019) find that distributing maternal and child health materials increased fathers’ support for mothers of newborns (in Indonesia). Thus, parenting interventions that do not target fathers explicitly can—but do not always—change their behavior and increase their level of responsive parenting. Taken together, these studies provide initial evidence that well-crafted parenting programs targeting fathers can increase their maternal and child health knowledge, change their behaviors, and lead to improved outcomes for mothers and children. Evans and Jakiela 219 However, there are potential pitfalls. Most notably, fathers appear less inclined to par- ticipate in parenting education programs, particularly those that are mainly targeted to mothers; this can limit the potential impacts on fathers. Moreover, there is some risk that parenting interventions can have unintended negative impacts on fathers, possi- bly because they lead fathers to attend to aspects of their interactions with children that they had previously accepted uncritically. In general, parenting interventions ap- pear more effective at improving child development when they target mothers (only), Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 with the potential for spillovers onto fathers’ practice, as opposed to fathers (only), but interventions that target both parents may be the most effective in some settings. These initial conclusions draw on the limited literature available to date and may change as the body of evidence expands. Other Impacts of ECD Interventions on Fathers Economic theory suggests that many ECD interventions, particularly center-based childcare and early childhood education, have obvious, direct implications for parental income and labor supply. Institutional childcare interventions have the potential to be win-win policies if they improve children’s developmental outcomes while also in- creasing parents’ labor force participation and household income. Furthermore, they may have indirect impacts on parenting activities through other channels: e.g., fathers who work more may have less time with their children, which could subsequently affect child development outcomes. Nevertheless, many impact evaluations of day- care, preschool, and kindergarten programs in LMICs do not even report impacts on mothers—though most of those that do find positive impacts on maternal labor force participation (Evans, Jakiela, and Knauer 2021; Halim, Perova, and Reynolds 2022).6 Much less evidence exists documenting the impacts—or lack of impacts—of institu- tional childcare and early childhood education on fathers’ labor supply and earned in- come. In this section, we outline the impacts of non-parenting ECD interventions on parenting practices and on other aspects of fathers’ lives. To date, only one study estimates the impact of preschool on fathers’ parenting prac- tices in LMICs. Lassassi (2021) uses a difference-in-differences strategy to estimate the impact of preschool availability on fathers’ interactions with their young children in Algeria, and does not find a statistically significant impact of preschool on fathers’ parenting. However, preschool is also associated with a decline in the quality of in- teractions between mothers and young children in that context.7 Likewise, only one randomized trial in an LMIC reports estimates of the impact of access to childcare on fathers’ labor market participation. Bjorvatn et al. (2022) estimate the impacts of offer- ing Ugandan households with young children (aged 3 to 5 at baseline) either subsidized childcare, a cash grant of equivalent value, or the combination of the two interven- tions. Both the childcare subsidy alone and the combination of subsidized childcare with a cash grant increased the likelihood that a young child is in full-time childcare 220 The World Bank Research Observer, vol. 40, no. 2 (2025) by almost fifty percentage points. Access to childcare increased fathers’ labor supply and earned income, effects which were driven by increases in wage labor rather than self-employment. In contrast, access to childcare alone did not increase mothers’ la- bor supply; though the likelihood of self-employment did increase for mothers who re- ceived both a childcare subsidy and a cash grant. Access to childcare also had positive and statistically significant impacts on both child development and mothers’ wellbe- ing, suggesting that it is indeed a win-win from a policy perspective in that context. Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 The number of quasi-experimental studies estimating the impact of daycare, preschool, and kindergarten on fathers labor market participation in LMICs is vanish- ingly small, making it difficult to draw broad conclusions from the literature. Rosero and Oosterbeek (2011) use a regression discontinuity design to estimate the impact of access to free center-based childcare on households in Ecuador. They find a posi- tive, statistically significant impact of childcare on the income of the household head, who was typically the father, as well as the income of the mother. However, in their set- ting, institutional childcare had null or even negative impacts on child development, suggesting a tradeoff between household income and children’s wellbeing. In contrast, Wang and Lin (2019) use a difference-in-differences strategy to estimate the impact of affordable preschool in China. They find that the expansion of childcare access in- creased rates of entrepreneurship among mothers, but not among fathers. In related work, Ohrnberger et al. (2020) find that child grants in South Africa increased fathers’ mental health. Selection into our sample of published studies is, of course, a concern, but these re- sults suggest that early childhood interventions can and do impact fathers by changing their parenting skills and practices, their involvement in the labor market, or poten- tially both. Impact evaluations that ignore the role that fathers play in raising children and balancing responsibilities inside and outside the home risk understating the true impacts of ECD interventions. Overall, it remains difficult to draw conclusions about the impacts of ECD interven- tions on fathers’ lives beyond parenting because the number of studies reporting such impacts is still so small. However, given the evidence that childcare interventions can impact fathers’ income (as in Rosero and Oosterbeek 2011 and Bjorvatn et al. 2022, discussed above), it is possible that documented impacts of institutional childcare on household income (Rosero and Oosterbeek 2011; Attanasio et al. 2022; Bjorvatn et al. 2022) might also be partially explained by impacts on fathers’ labor supply and wages. This highlights the importance of estimating the impacts of ECD interventions on all household members, and not just children and mothers. Discussion One of the main takeaways from our survey of empirical research on fathers’ in- volvement in parenting young children in LMICs is that the evidence base remains Evans and Jakiela 221 quite small. Given this, it is possible that conclusions will change as the number of well-identified studies expands. All the studies that we consider are of relatively high quality, but there is some variation in the potential risk of bias. We discuss study quality in Online Appendix A2. Most of the studies that we discuss are random- ized trials with levels of attrition that are either low or well-balanced across treat- ment arms—and we do not find evidence that patterns of results differ between ex- perimental and quasi-experimental studies, or between studies with lower vs. higher Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 attrition. We focus on the role that fathers play in parenting young children in LMICs. Mothers also do more active parenting than fathers in wealthy countries. For example, Guryan et al. (2008) show that fathers spend about half as much time providing childcare as mothers in the United States (6.8 hours per week versus 14 hours), with similar ra- tios between men and women in Canada, the Netherlands, and Norway and much less equal ratios in some higher income countries (France and Austria). In the United States, at least, the amount of time that fathers spend on childcare has been increasing rapidly over time (Doepke and Zilibotti 2017). However, parenting norms, educational systems, and labor markets in industrial- ized countries differ from those found in lower income settings. This may explain why the correlates of parental involvement appear different in high-income countries, where systematic reviews point to strong positive associations between paternal pres- ence and engagement and child outcomes. Sarkadi et al. (2007) review 24 longitudi- nal studies and find mostly positive associations between paternal involvement and child development in high-income country environments, including fathers’ presence (i.e., cohabitation with the child) and fathers’ engagement with their children. Posi- tive child development outcomes include fewer behavioral problems among boys and fewer psychological problems among young women, as well as reduced criminal be- havior among youth in lower income households. Some studies also suggest a posi- tive link between father-child play and children’s cognitive socio-emotional outcomes in high-income settings (Amodia-Bidakowska, Laverty, and Ramchandani 2020). As LMICs continue to industrialize, it remains to be seen whether the empirical patterns associated with fathers’ active involvement in parenting will become more similar to those observed in higher income countries. Alloparenting is another way that childrearing practices in LMICs differ system- atically from those observed in industrialized nations. In many LMIC settings, older siblings—particularly sisters—and grandmothers play a major role in childrearing (Lancy 2015; Evans, Jakiela, and Knauer 2021). To date, few studies document the im- pacts of ECD interventions on older siblings or grandparents. Alsan (2017) shows that a vaccination campaign targeted to children under five years old in Turkey boosted literacy and educational attainment among older sisters but not older brothers, likely because older sisters were more likely to stay home and care for ill siblings. Consis- 222 The World Bank Research Observer, vol. 40, no. 2 (2025) tent with that hypothesis, the effects were larger in households where mothers worked outside of the home. The provision of community-led preschools in Mozambique re- duced the amount of time older siblings spent caring for young children and increased their time spent on schoolwork, though it did not have a statistically significant im- pact on enrollment (Martinez, Naudeau, and Pereira 2017). In urban Brazil, Attanasio et al. (2022) find that access to government-run daycare centers boosted labor mar- ket outcomes among both grandmothers and siblings aged 15 or older. Though active, Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 responsive parenting may be optimal from a child development perspective, parents in LMICs are often engaged in pressing tasks such as income generation and subsis- tence farming, which is why they commonly assign children in middle childhood and adolescence the task of caring for their younger (below school-aged) siblings (Weisner et al. 1977, Hrdy 2009). Given this, the limited evidence on the role that older children play in the lives of their siblings in early childhood seems like an important gap in the literature, just like the role of fathers. Conclusion Fathers are parents too, but their role in parenting—particularly in LMIC contexts— is often ignored. Historically, fathers have played a less important role in parenting than mothers, but this gap is narrowing as societies develop. As women become more educated and enter the workforce, the traditional division of household responsibilities is evolving, and fathers are likely to play a growing role in the lives of young children in the future than they have in the past. The existing evidence base on fathers’ roles in early childhood development is small, but it is expanding rapidly. It is increasingly clear that well-designed, contextually- appropriate interventions targeting fathers—or fathers and mothers together—can change fathers’ parenting knowledge in LMIC settings. It remains to be seen whether targeting fathers improves child and household outcomes. Indeed, a review of inter- ventions intended to engage fathers showed that most apply the same program de- signs to fathers that have been applied to mothers, with little adaptation (Jeong et al. 2023b). Yet evidence suggests that fathers face specific challenges, such as less time at home, less knowledge about caregiving, and restrictive gender attitudes (Jeong et al. 2023c). Fathers may be less inclined to participate in parenting classes, and they may be less likely than mothers to put their newfound parenting knowledge into prac- tice. Moreover, fathers can also be impacted by interventions that do not target them directly, particularly when access to institutional childcare changes the labor supply decisions facing the household. Understanding when and how to target fathers to max- imize child development and household welfare gains requires an expansion of the evidence base documenting the impacts of all types of ECD interventions on fathers’ behavior. Evans and Jakiela 223 Data Availability Statement Stata code used to produce Figure 1 from the publicly available MICS data is posted on the authors’ websites. Notes Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 David K. Evans, Inter-American Development Bank; email address is devans@iadb.org; Pamela Jakiela (corresponding author) Williams College, BREAD, CGD, IPA, IZA, and J-PAL; email address is pj5@williams.edu. The authors thank Echidna Giving, the Hewlett Foundation, and the World Bank for funding. Authors are listed alphabetically. Amina Mendez Acosta provided extensive research assistance. Heather Knauer, Megan O’Donnell, Owen Ozier, and Elizaveta Perova provided helpful comments. All errors are our own. 1. Data from the World Bank’s World Development Indicators database indicates that gross prepri- mary enrollment in LMICs rose from 29 percent in 2001 to 58 percent in 2018. Pupil-teacher ratios in preprimary remained relatively constant over the same period, suggesting that levels of government in- vestment in early childhood education were also rising over time. World Bank funding for early childhood education programs has also increased substantially over the last two decades (Sayre et al. 2015). 2. MICS have been conducted in 118 countries, but not all surveys measure early childhood stimula- tion. Appendix Table A3.1 lists the surveys we analyzed, which are drawn from the three most recent MICS rounds (4, 5, and 6) and cover the period from 2010–2023. 3. The share of children living with their mother varies considerably less, from 85 percent in Sub- Saharan Africa to 99 percent in the Middle East and North Africa. 4. Another study—not included in our sample because it ultimately did not estimate impacts on fathers—initially included a treatment arm that encouraged men to participate in parent training classes in rural Mexico, but take-up was so negligible that the evaluation of father training was abandoned (Cárdenas, Evans, and Holland 2023). 5. However, in contrast to Justino et al. (2023), Luoto et al. (2021) find suggestive evidence that the program did improve the parenting practices of fathers who participated, and Abimpaye et al. (2020) find that the same program increased fathers’ involvement in early childhood stimulation. 6. Rosero and Oosterbeek (2011); Martinez, Naudeau, and Pereira (2017); Hojman and López Boo (2019) and Clark et al. (2019) all find that access to childcare (either daycare or preschool) increases ma- ternal labor supply, while Attanasio et al. (2022) and Richardson et al. (2018) do not. Attanasio et al. (2022) find that access to daycare increases labor force participation among adult sisters and grandmothers; in their context, most mothers are already working, and their labor supply is not affected by access to child- care. 7. Two other studies that came up in our systematic review find that nutrition interventions (without a cash grant component) increase male income or labor supply (Fitzsimons et al. 2016, Adams et al. 2018); this result is surprising, since child nutrition programs do not inherently free up parental time. References Abimpaye, M., C. Dusabe, J. P. Nzabonimpa, R. Ashford, and L. Pisani. 2020. “Improving Parenting Practices and Development for Young Children in Rwanda: Results Form a Randomized Control Trial.” Interna- tional Journal of Behavioral Development 44(3): 205–15. Adams, K. P., T. J. Lybbert, S. A. Vosti, E. Ayifah, M. Arimond, S. Adu-Afarwuah, and K. G. Dewey. 2018. “Un- intended Effects of a Targeted Maternal and Child Nutrition Intervenient on Household Expenditures, 224 The World Bank Research Observer, vol. 40, no. 2 (2025) Labor Income, and the Nutritional Status of Non-Targeting Siblings in Ghana.” World Development 107: 138–15. Akresh, R. 2009. “Flexibility of Household Structure: Child Fostering Decisions in Burkina Faso.” Journal of Human Resources 44(4): 976–99. Alsan, M. 2017. “The Gendered Spillover Effect of Young Children’s Health on Human Capital: Evidence from Turkey.” NBER Working Paper No. 2370, National Bureau of Economic Research, Cambridge, MA. Amaral, S., L. Dinarte, P. Dominguez, and S. M. Perez-Vincent. 2021. “Helping Families Help Themselves? Heterogeneous Effects of a Digital Parenting Program.” Policy Research Working Paper No. 9850, World Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 Bank, Washington, DC. Amodia-Bidakowska, A., C. Laverty, and P. G. Ramchandani. 2020. “Father-Child Play: A Systematic Review of Its Frequency, Characteristics and Potential Impact on Children’s Development.” Developmental Re- view 57: 100924. Antelman, G., J. Ferla, M. M. Gill, H. J. Hoffman, T. Komba, A. Abubakar, P. Remes, O. Jahanpour, M. Mariki, M. A. Mang’enya, and R. van de Ven 2023. “Effectiveness of an Integrated Multilevel Early Child Devel- opment Intervention on Caregiver Knowledge and Behavior: A Quasi-experimental Evaluation of the Malezi Program in Tanzania.” BMC Public Health 23(1): 19. Attanasio, O., R. Paes de Barros, P. Carneiro, D. K. Evans, L. Lima, P. Olinto, and N. Schady. 2022. “Pub- lic Childcare, Labor Market Outcomes of Caregivers, and Child Development: Experimental Evidence from Brazil.” NBER Working Paper No. 30653, National Bureau of Economic Research, Cambridge, MA. Becker, G. S. 1981. A Treatise on the Family. Cambridge, MA: Harvard University Press. Beegle, K., D. Filmer, A. Stokes, and L. Tiererova. 2010. “Orphanhood and the Living Arrangements of Chil- dren in Sub-Saharan Africa.” World Development 38(12): 1727–46. Björkman Nyqvist, M., and S. Jaychandran. 2017. “Mothers Care More, but Fathers Decide: Educating Par- ents about Child Health in Uganda.” American Economic Review: Papers and Proceedings 107(5): 496– 500. Bjorvatn, K., D. Ferris, S. Gulesci, A. Nasgowitz, V. Somville, and L. Vandewalle. 2022. “Childcare, Labor Supply, and Business Development: Experimental Evidence from Uganda.” G2 LM|LIC Working Paper No. 67. Black, M. M., S. P. Walker, L. C. H. Fernald, C. T. Andersen, A. M. DiGirolamo, C. Lu, D. C. McCoy, G. Fink, Y. R. Shawar, J. Shiffman, A. E. Devercelli, Q. T. Wodon, E. Vargas-Baron, and S. Grantham-McGregorand the Lancet Early Childhood Development Series Steering Committee 2017. “Early Childhood Development Coming of Age: Science through the Life Course.” Lancet 389(10064): 77–90. Bos, J. M., A. Khan, S. Ravindran, and A. Shonchoy. 2024. “Early Childhood Human Capital Formation at Scale.” Journal of Public Economics 231: 105046. Boserup, E. 1970. Woman’s Role in Economic Development. Abingdon, UK: Earthscan. Britto, P. R., S. J. Lye, K. Proulx, A. K. Yousafzai, S. G. Matthews, T. Vaivada, R. Perez-Escamilla, N. Rao, P. Ip, L. C. H. Fernald, H. MacMillan, M. Hanson, T. D. Wachs, H. Yao, H. Yoshikawa, A. Cerezo, J. F. Leckman, and Z. A. Bhuttaand the Early Childhood Development Interventions Review Group for the Lancet Early Childhood Development Series Steering Committee 2017. “Nurturing Care: Promoting Early Childhood Development.” Lancet 389(10064): P91–102. Cárdenas, S., D. K. Evans, and P. Holland. 2023. “Parent Training and Child Development at Low Cost? Evidence from a Randomized Field Experiment in Mexico.” Journal of Research in Childhood Educa- tion 38(sup 1): S130–S160. Carneiro, P., L. Kraftman, G. Mason, L. Moore, I. Rasul, and M. Scott. 2021. “The Impacts of a Multi- faceted Prenatal Intervention on Human Capital Accumulation in Early Life.” American Economic Re- view 111(8): 2506–49. Case, A., and C. Ardington. 2006. “The Impact of Parental Death on School Outcomes: Longitudinal Evi- dence from South Africa.” Demography 43(1): 401–20. Evans and Jakiela 225 Clark, S., C. W. Kabiru, S. Laszlo, and S. Muthuri. 2019. “The Impact of Childcare on Poor Urban Women’s Economic Empowerment in Africa.” Demography 56(4): 1247–72. Cuartas, J., J. Jeong, C. Rey-Guerra, D. C. McCoy, and H. Yoshikawa. 2020. “Maternal, Paternal, and Other Caregivers’ stimulation in Low-and-middle-income Countries.” PLoS ONE 15(7): e0236107. Davis, J., and N. Brazil, 2016. Migration, remittances and nutrition outcomes of left-behind children: a national-level quantitative assessment of Guatemala. PloS One 11(3): e0152089. Dinga, L. A. 2019. “Effect of Father-Targeted Nutrition Education on Feeding Practices, Nutritional Status and Morbidity among Infants in Kisumu East, Kenya”. Doctoral thesis, Jomo Kenyatta University of Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 Agriculture and Technology. Doepke, M., and F. Zilibotti. 2017. “Parenting with Style: Altruism and Paternalism in Intergenerational Preference Transmission.” Econometrica 85(5): 1331–71. Doyle, K., R. G. Levtov, G. Barker, G. G. Bastian, J. B. Bingenheimer, S. Kazimbaya, A. Nzabonimpa, J. Puler- witz, F. Sayinzoga, V. Sharma, and D. Shattuck 2018. “Gender-Transformative Bandebereho Couples’ intervention to Promote Male Engagement in Reproductive and Maternal Health and Violence Pre- vention in Rwanda: Findings from a Randomized Controlled Trial.” PLoS ONE 13(4): e0192756. Engle, P. L., L. C. H. Fernald, H. Alderman, J. Behrman, C. O’Gara, A. Yousafzai, M. C. D. de Mello, M. Hidrobo, N. Ulkuer, I. Ertem, and S. Iltus, and the Global Child Development Steering Group 2011. “Strategies for Reducing Inequalities and Improving Developmental Outcomes for Young Children in Low-Income and Middle-Income Countries.” Lancet 378(9799): 1339–53. Evans, D. K., P. Jakiela, and H. A. Knauer. 2021. “The Impact of Early Childhood Interventions on Mothers.” Science 372(6544): 794–6. Evans, D. K., and E. Miguel. 2007. “Orphans and Schooling in Africa: A Longitudinal Analysis.” Demography 44(1): 35–57. Evans, D.K., P. Jakiela, and A. Mendez Acosta, 2024. "The Impacts of Childcare Interventions on Children’s Outcomes in Low- and Middle-Income Countries: A Systematic Review." AEA Papers and Proceedings 114: 463–6. Fitzsimons, E., B. Malde, A. Mesnard, and M. Vera-Hernández. 2016. “Nutrition, Information and Household Behavior: Experimental Evidence from Malawi.” Journal of Development Economics 122: 113–26. Gertler, P., J. Heckman, R. Pinto, A. Zanolini, C. Vermeersch, S. Walker, S. M. Chang, and S. Grantham- McGregor. 2014. “Labor Market Returns to an Early Childhood Stimulation Intervention in Jamaica.” Science 344(6187): 998–1001. Guryan, J., E. Hurst, and M. Kearney. 2008. “Parental Education and Parental Time with Children.” Journal of Economic Perspectives 22(3): 23–46. Halim, D., E. Perova, and S. Reynolds. 2022. “Childcare and Mothers’ Labor Market Outcomes in Lower- and Middle-Income Countries.” World Bank Research Observer 38(1): 73–114. Hamadani, J. D., F. Tofail, A. Hilaly, S. N. Huda, P. Engle, and S. M. Grantham-McGregor. 2010. “Use of Fam- ily Care Indicators and Their Relationship with Child Development in Bangladesh.” Journal of Health Population and Nutrition 28(1): 23–33. Hojman, A., and F. López Boo. 2019. “Cost-Effective Public Daycare in a Low-Income Economy Benefits Children and Mothers.” Working Paper Series No. 1036, Inter-American Development Bank. Hrdy, S. 2009. Mothers and Others: The Evolutionary Origins of Mutual Understanding. Cambridge, MA: Belk- nap Press. Jensen, S. K. G., M. Placencio-Castro, S. M. Murray, R. T. Brenna, S. Goshev, J. Farrar, A. Yousafzai, L. B. Rawl- ings, B. Wilson, E. Habyarimana, V. Sezibera, and T. Betancourt 2021. “Effect of a Home-Visiting Parent- ing Program to Promote Early Childhood Development and Prevent Violence: A Cluster-Randomized Trial in Rwanda.” BMJ Global Health 6: e003508. 226 The World Bank Research Observer, vol. 40, no. 2 (2025) Jeong, J., D. C. McCoy, A. K. Yousafzai, C. Salhi, and G. Fink. 2016. “Paternal Stimulation and Early Child Development in Low-and Middle-Income Countries.” Pediatrics 138(4): e20161357 Jeong, J., E. F. Sullivan, and J. K. McCann. 2023a. “Effectiveness of Father-Inclusive Interventions on Mater- nal, Paternal, Couples, and Early Child Outcomes in Low- and Middle-Income Countries: A Systematic Review.” Social Science & Medicine 328: 115971. Jeong, J., E. F. Sullivan, J. K. McCann, D. C. McCoy, and A. K. Yousafzai. 2023b. “Implementation Character- istics of Father-Inclusive Interventions in Low- and Middle-Income Countries: A Systematic Review.” Annals of the New York Academy of Sciences 1520(1): 34–52. Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 Jeong, J., J. K. McCann, A. Alsager, A. Bhojani, N. Andrew, J. Joseph, M. N. Ahun, M. Kabati, and D. Joachim. 2023c. “Formative Research to Inform the Future Design of a Multicomponent Fatherhood Intervention to Improve Early Child Development in Mwanza, Tanzania.” Social Science & Medicine 331: 116072. Jukes, M., P. Gabrieli, N. L. Mgonda, F. Nsolezi, G. Jeremiah, J. Tibenda, and K. L. Bub. 2018. “Respect Is an Investment: Community Perceptions of Social and Emotional Competencies in Early Childhood from Mtwara, Tanzania.” Global Education Review 5(2): 160–88. Justino, P., M. Leone, P. Rolla, M. Abimpaye, C. Dusabe, M. D. Uwamahoro, and R. Germond 2023. “Improv- ing Parenting Practices for Early Child Development: Experimental Evidence from Rwanda.” Journal of the European Economic Association 21(4): 1510–50. Kariger, P., E. A. Frongillo, P. Engle, P. M. R. Britto, S. M. Sywulka, and P. Menon. 2012. “Indicators of Fam- ily Care for Development for Use in Multicountry Surveys.” Journal of Health Population and Nutrition 30(4): 472–86. Lancy, D. F. 2015. The Anthropology of Childhood: Cherubs, Chattel, Changelings(2nd edition). Cambridge, UK: Cambridge University Press. Lassassi, M. 2021. “Does Preschool Improve Child Development and Affect the Quality of Parent-child In- teraction? Evidence from Algeria.” International Journal of Educational Development 82: 102354. Li, Q., G. Liu, and W. Zang. 2015. “The Health of Left-Behind Children in Rural China.” China Economic Review 36: 367–76. Luoto, J. E, I. Lopez Garcia, F. E. Aboud, D. R. Singla, L. C. H. Fernald, H. O. Pitchik, U. Y. Saya, R. Otieno, and E. Alu 2021. “Group-based Parenting Interventions to Promote Child Development in Rural Kenya: A Multi-Arm, Cluster-Randomized Community Effectiveness Trial.” Lancet Global Health 9: e309–19. Mansuri, G. 2006. “Migration, School Attainment, and Child Labor: Evidence from Rural Pakistan.” Policy Research Working Paper No. 3945, World Bank, Washington, DC. Martinez, S., S. Naudeau, and V. A. Pereira. 2017. “Preschool and Child Development under Extreme Poverty: Evidence from a Randomized Experiment in Rural Mozambique.” Policy Research Working Paper No. 8290, World Bank, Washington, DC. Munshi, K. 2020. “Social Networks and Migration.” Annual Review of Economics 12: 503–24. Ohrnberger, J., E. Fichera, M. Sutton, and L. Anselmi. 2020. “The Effect of Cash Transfers on Mental Health – New Evidence from South Africa.” BMC Public Health [Electronic Resource] 20(article number 436): https://doi.org/10.1186/s12889- 020- 08596- 7. Osaki, K., T. Hattori, A. Toda, E. Mulati, L. Hermawan, K. Pritasari, S. Bardosono, and S. Kosen 2019. “Ma- ternal and Child Health Handbook Use for Maternal and Child Care: A Cluster Randomized Controlled Study in Rural Java, Indonesia.” Journal of Public Health 41(1): 170–82. Özlüses, E., and A. Çelebioglu. 2014. “Educating Fathers to Improve Breastfeeding Rates and Paternal- Infant Attachment.” Indian Pediatrics 51: 654–7. Pitt, M. M., M. R. Rosenzweig, and M. N. Hassan. 2012. “Human Capital Investment and the Gender Division of Labor in a Brawn-Based Economy.” American Economic Review 102(7): 3531–60. Rahman, A., A. Malik, S. Sikander, C. Roberts, and F. Creed. 2008. “Cognitive Behavior Therapy-Based In- tervention by Community Health Workers for Mothers with Depression and Their Infants in Rural Pak- istan: A Cluster-randomized Controlled Trial.” Lancet 372: 902–9. Evans and Jakiela 227 Richardson, H., G. Lisandrelli, A. Riobueno-Naylor, and R. Saxe. 2018. “Development of the Social Brain from Age Three to Twelve Years.” Nature Communications 9(1027): 1–12. Rosero, J., and H. Oosterbeek. 2011. “Trade-offs between Different Early Childhood Interventions: Evidence from Ecuador.” Tinbergen Institute Discussion Paper TI 2011-102/3. Sarkadi, A., R. Kristiansson, F. Oberklaid, and S. Bremberg. 2007. “Fathers’ Involvement and Children’s De- velopmental Outcomes: A Systematic Review of Longitudinal Studies.” Acta Paediatrica 97(2): 153–8. Sayre, R. K., A. E. Devercelli, M. J. Neuman, and Q. Wodon. 2015. Investing in Early Childhood Development: Review of the World Bank’s Recent Experience. Washington, DC: World Bank. Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 Schmeer, K. 2009. “Father Absence due to Migration and Child Illness in Rural Mexico.” Social Science & Medicine 69(8): 1281–6. Sear, R., and R. Mace. 2008. “Who Keeps Children Alive? A Review of the Effects of Kin on Child Survival.” Evolution and Human Behavior 29(1): 1–18. Wang, Q., and M. Lin. 2019. “Work-Family Policy and Female Entrepreneurship: Evidence from China’s Subsidized Childcare Program.” China Economic Review 54: 256–70. Wang, Y., Z. Huang, M. Yang, F. Wang, and S. Xiao. 2015. “Reducing Environmental Tobacco Smoke Ex- posure of Preschool Children: A Randomized Controlled Trial of Class-Based Health Education and Smoking Cessation Counseling for Caregivers.” International Journal of Environmental Research and Public Health 12(1): 692–709. Weber, A., A. Fernald, and Y. Diop. 2017. “When Cultural Norms Discourage Talking to Babies: Effectiveness of a Parenting Program in Rural Senegal.” Child Development 88(5): 1513–26. Weisner, T. S., R. Gallimore, M. K. Bacon, H. Barry, III, C. Bell, S. C. Novaes, C. P. Edwards, B. B. Goswami, L. Minturn, S. Nerlove, A. Koel, J. E. Ritchie, P.C. Rosenblatt, T. R. Singh, B. Sutto-Smith, B. B. Whiting, W. D. Wilder, and T. R. Williams 1977. “My Brother’s Keeper: Child and Sibling Caretaking.” Current Anthropology 18(2): 169–90. Yabiku, S. T., V. Agadjanian, and B. Cau. 2012. “Labor Migration and Child Mortality in Mozambique.” Social Science & Medicine 75(12): 2530–8. Zhao, Q., X. Yu, X. Wang, and T. Glauben. 2014. “The Impact of Parental Migration on Children’s School Performance in Rural China.” China Economic Review 31: 43–54. 228 The World Bank Research Observer, vol. 40, no. 2 (2025) Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 Supplementary Online Appendix The Role of Fathers in Promoting Early Childhood Development in Low- and Middle-Income Countries: A Review of the Evidence David K. Evans and Pamela Jakiela Appendix A1: Search for Studies ECD Interventions in LMICs That Report Father-Specific Outcomes We initially carried out a systematic search for articles published between 2005 and 2019 in EconLit, Pubmed, Web of Science, and PsychINFO. We used a variety of search terms to identify evaluations of early childhood development interventions (i.e., tar- geted at children ages 0–5 or their caregivers): e.g., “early child” or “early childhood” Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 or “infant” or “young children” or “birth” or “early childcare.” We also included search terms for broad classes of interventions (e.g., cash transfer or daycare or food supple- mentation or home visit). We included studies that evaluated interventions in low- and middle-income countries, and we included both experimental and quasi-experimental evaluations. Through this process, we initially identified 3,716 unique, potential stud- ies. We then excluded studies that appeared in the search but did not satisfy the inclu- sion criteria above and added studies that the authors were aware of and did not appear in the search. This resulted in 478 experimental and quasi-experimental evaluations of ECD interventions in LMICs. Of those, only 12 reported any outcomes specific to the father or male head-of-household. That initial search was completed in the end of 2019. Exhaustive details of that initial search are included in the supplementary materials of Evans, Jakiela, and Knauer (2021). We then systematically updated our sample by examining all studies that cite those 12 papers and applying the same inclusion criteria described above. This update was finalized in June 2022. This resulted in the inclusion of 10 additional studies, resulting in a total sample of 22 studies. When we became aware that a study included in our sample had been published, we checked the results for consistency with the earlier draft and updated the reference, which explains why some of the studies in our sample have publication dates after 2022. Appendix A2: Study Quality Not all the studies we include are of equal quality. There are various factors that con- tribute to study quality. We use just two factors to provide a simple classification. The first is study design. Because we exclude study designs most likely to incorporate omit- ted variable bias from our review (propensity score matching or simple multivariate regression), all the study designs are relatively high quality. We propose that random- ized controlled trials (RCTs) have the lowest probability of omitted variable bias, so we rank those as high. We also examine study attrition: studies with lower than 20 percent with relatively good balance (within 5 percentage points) across groups are considered low attrition. None of our studies have attrition higher than 35 percent, so again, we exclude studies with greatest risk of attrition bias. We rate RCTs with low, balanced attrition as high quality, RCTs with higher or slightly imbalanced attrition as medium-high, quasi-experimental studies with low attrition as medium-high, and quasi-experimental studies with higher attrition as medium. We treat a lack of attrition analysis the same as high attrition. This analysis leaves us with, out of a total of 22 studies, 12 high quality studies, 5 medium-high quality studies, and 5 medium quality studies. Because we include the underlying values in Appendix Table A3.3 below, readers can use their own weights across studies. Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 Appendix A3 Table A3.1. MICS Surveys Used in Analysis Country MICS Round Survey Year(s) Afghanistan 6 2022–2023 Algeria 6 2018–2019 Argentina 6 2019–2020 Bangladesh 6 2019 Barbados 4 2012 Belarus 6 2019 Belize 5 2015–2016 Benin 5 2014 Bhutan 4 2010 Bosnia and Herzegovina 4 2011–2012 Cameroon 5 2014 Central African Republic 6 2018–2019 Chad 6 2019 Comoros 6 2022 Congo, Dem. Rep. 6 2017–2018 Congo, Rep. 5 2014–2015 Costa Rica 6 2018 Cote d’Ivoire 5 2016 Cuba 5 2014 Dominican Republic 5 2014 El Salvador 5 2014 Eswatini 6 2021–2022 Fiji 6 2021 Georgia 6 2018 Ghana 6 2017–2018 Guinea 5 2016 Guinea-Bissau 6 2018–2019 Guyana 5 2014 Evans and Jakiela 1 Table A3.1. Continued Country MICS Round Survey Year(s) Honduras 6 2019 Iraq 6 2018 Jamaica 4 2011 Kazakhstan 5 2015 Kiribati 6 2018–2019 Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 Kyrgyz Republic 5 2014 Lao PDR 4 2011–2012 Lesotho 6 2018 Malawi 6 2019–2022 Mali 5 2015 Mauritania 5 2015 Mexico 5 2015 Moldova 4 2012 Mongolia 6 2018 Montenegro 6 2018 Nepal 6 2019 Nigeria 4 2011 Palestine 6 2019–2020 Panama 5 2013 Paraguay 5 2016 Qatar 4 2012 Saint Lucia 4 2012 Sao Tome and Principe 6 2019 Samoa 6 2019–2020 Serbia 6 2019 Sierra Leone 6 2017 Suriname 6 2018 Thailand 4 2012–2013 The Gambia 6 2018 Togo 6 2017 Tonga 6 2019 Trinidad and Tobago 4 2011 Tunisia 4 2011–2012 Turkmenistan 5 2015–2016 Turks and Caicos 6 2019–2020 Tuvalu 6 2019–2020 Ukraine 4 2012 Uruguay 4 2012–2013 Uzbekistan 6 2021–2022 Vietnam 4 2010–2011 Zimbabwe 6 2019 Source: Authors’ analysis. Note: We analyze the most recent survey for each country included in rounds 4, 5, and 6 of the MICS, covering the period 2010-2015. Table A3.2. Impact Evaluations of ECD Interventions in LMICs That Report Father-Specific Outcomes Study Country Intervention Design Father Outcomes Reported Coef. SE/CI/p-value Abimpaye et al. (2020) Rwanda parenting education RCT Parenting outcome: father learning/play activities Treatment: light touch intervention 0.51 p < 0.001 Treatment: full intervention 0.67 p < 0.001 Parenting outcome: father nurturing care activities Treatment: light touch intervention 0.33 p < 0.001 Treatment: full intervention 0.31 p < 0.001 Parenting outcome: father harsh discipline Treatment: light touch intervention 0.06 p > 0.05 Treatment: full intervention 0.05 p > 0.05 Adams et al. (2018) Ghana lipid-based nutrient, iron, and folic RCT Mother’s husband’s income 0.109 (0.055) acid supplements during pregnancy and post-partum Amaral et al. (2021) El Salvador digital stress management and RCT Parenting outcomes: positive parenting intervention Positive caregiver-child interactions −0.137 (0.067) Violence in parenting (physical and 0.025 (0.065) psychological) Father’s tolerance of violent parenting 0.019 (0.066) Other father outcomes: Father’s mental distress (DASS-21) 0.093 (0.05) Father’s impulsiveness (Barratt Impulsiveness −0.074 (0.062) Scale BIS-11) Antelman et al. (2023) Tanzania Community health workers DD Father engagement (adjusted odds ratio) 1.90 [1.45,2.48] showing ECD videos to caregivers Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 Table A3.2. Continued Study Country Intervention Design Father Outcomes Reported Coef. SE/CI/p-value Björkman Nyqvist and Uganda health and nutrition classes RCT Parenting outcome: father’s health knowledge Jayachandran (2017) Treatment: men’s parenting classes 0.22 (0.045) Treatment: women’s classes (spillover to fathers) 0.018 (0.042) Parenting outcome: household health behaviors Treatment: men’s parenting classes 0.088 (0.064) Bjorvatn et al. (2022) Uganda preschool vouchers, and an RCT Outcome: father’s self-employment income equivalent cash grant, or combination of both interventions Treatment: Childcare only 2.61 (4.03) Treatment: Cash only −5.49 (3.56) Treatment: Childcare and cash 1.87 (3.84) Outcome: father’s income from wage labor Treatment: childcare 18.1 (9.12) Treatment: cash equivalent 8.02 (8.97) Treatment: childcare plus cash equivalent 0.92 (8.82) Outcome: father’s total earned income Treatment: childcare 24.56 (10.16) Treatment: cash equivalent 5.08 (9.88) Treatment: childcare plus cash equivalent 5.04 (9.94) Outcome: father’s self-employment hours worked Treatment: childcare −3.38 (8.21) Treatment: cash equivalent −0.4 (8.54) Treatment: childcare plus cash equivalent 10.43 (8.76) Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 Table A3.2. Continued Study Country Intervention Design Father Outcomes Reported Coef. SE/CI/p-value Outcome: father’s wage labor hours worked Treatment: childcare 20.5 (9.55) Treatment: cash equivalent 8.06 (9.09) Treatment: childcare plus cash equivalent 7.56 (9.33) Outcome: father’s total hours worked Treatment: childcare 18.29 (11.68) Treatment: cash equivalent 8.41 (11.68) Treatment: childcare plus cash equivalent 16.27 (11.79) Outcome: value of father’s business assets (1000s of Ugandan shillings) Treatment: childcare 0.98 −1.13 Treatment: cash equivalent 1.97 −1.31 Treatment: childcare plus cash equivalent 0.47 −1.02 Outcome: father’s employees Treatment: childcare 0.03 (0.05) Treatment: cash equivalent 0.03 (0.04) Treatment: childcare plus cash equivalent 0.06 (0.07) Bos et al. (2024) Bangladesh home-visiting intervention RCT Outcome: variety of learning activities provided −0.001 (0.076) providing educational materials by father and counseling Carneiro et al. (2021) Nigeria parenting education and cash RCT Parenting outcomes: transfers Father’s parenting knowledge (index) 0.257 (0.048) Other father outcomes: Outcome: any work in past year 0.003 (0.002) Outcome: days per week in highest-earning 0.394 (0.197) activity Outcome: self-employed 0.033 (0.021) Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 Table A3.2. Continued Study Country Intervention Design Father Outcomes Reported Coef. SE/CI/p-value Outcome: farms own land 0.001 (0.007) Outcome: monthly expenditure on father’s −4.83 −4.1 business Outcome: monthly earnings 16.7 (10.4) Dinga (2019) Kenya breastfeeding education for fathers RCT Outcome: knows breastmilk is first food 2.6 [0.11] of newborns Outcome: knows should start breastfeeding 21.8 [0.01] within 1 hour Outcome: knows should breastfeed for 2 years 15.9 [0.02] Outcome: knows about exclusive breastfeeding 23.8 [0.02] Doyle et al. (2018) Rwanda men’s groups, discussions of gender RCT Outcome: mother experienced physical violence 0.38 [0.29,0.50] equality (OR) Outcome: mother experienced sexual violence 0.36 [0.25,0.50] (OR) Outcome: use of physical punishment (OR) 0.66 [0.50,0.89] Outcome: sharing of tasks at home 0.33 [0.26,0.41] Outcome: time on household tasks 0.86 [0.49,1.23] Fitzsimons et al. (2016) Malawi nutrition education for parents RCT Outcome: male labor supply (index) 0.262 [0.131] through home visits Outcome: adult male works 0.106 [0.080] Outcome: adult male has two jobs 0.08 [0.025] Outcome: adult male’s work hours 4.314 [2.918] Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 Table A3.2. Continued Study Country Intervention Design Father Outcomes Reported Coef. SE/CI/p-value Jensen et al. (2021) Rwanda home-visits linked to Rwanda’s RCT Outcome: father engagement (OR) 1.592 [1.069, 2.368] social protection system Justino et al. (2023) Rwanda group-based parenting education RCT Outcome: father’s parenting time investment index (12 months post-treatment) Treatment: light treatment 0.177 (0.094) Treatment: full treatment 0.537 (0.216) Outcome: father’s parenting time investment index (33 months post-treatment) Treatment: light treatment 0.036 (0.08) Treatment: full treatment 0.183 (0.056) Outcome: father’s influence index (after 12 months) Treatment: light treatment 0.452 (0.225) Treatment: full treatment 0.681 (0.194) Lassassi (2021) Algeria preschool DD Outcome: fathers’ interactions with children 1.194 (1.579) Luoto et al. (2021) Kenya integrated responsive stimulation RCT Outcome: father’s Family Care Indicator behavioral score (0–6) and nutrition education Treatment: group-based parenting classes 0.08 [−0.10, 0.26] Treatment: group-based parenting classes plus 0.03 [−0.15, 0.22] home visits Ohrnberger et al. (2020) South unconditional cash transfer IV Outcome: effect on mental health of male adults 0.468 (0.447) Africa in the household (CES-D, higher score means better mental health) Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 Table A3.2. Continued Study Country Intervention Design Father Outcomes Reported Coef. SE/CI/p-value Osaki et al. (2019) Indonesia distribution of maternal and child RCT Outcome: husband’s support for saving for 1.82 [1.2,2.76] health handbooks delivery (OR) Outcome: husband’s support for keeping baby 1.58 [1.02,2.46] warm (OR) Outcome: husband’s support for child 1.62 [1.06,2.48] stimulation (OR) Özlüses and Çelebioglu Turkey breastfeeding education DD Outcome: paternal-infant attachment (2014) Treatment: education for mothers 9 NA Treatment: education for both parents 16.2 NA Rahman et al. (2008) Pakistan home visits RCT Outcome: father’s play frequency with infant at 1.9 [1.59, 4.15] 12 months Rosero and Oosterbeek Ecuador childcare centers, home visits RDD Outcome: household head income (2011) Treatment: home visits 5.466 (37.832) Treatment: childcare centers 97.621 (31.572) Wang et al. (2015) China smoking cessation counseling, RCT Outcome: father’s 7-day quit rate (OR) 1.12 [1.02,1.22] health education for children Wang and Lin (2019) China preschool DD Outcome: father’s entrepreneurship 0.025 (0.026) Source:Authors’ analysis. Note: The full references for all 22 of these studies are included in the main paper references. Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 Table A3.3. Study Quality Rating Attrition Study Design Overall rate? Balanced attrition? Overall quality Abimpaye et al. 2020 RCT 10% 2 to 5 p.p. higher attrition in Medium-high treatment arms Adams et al. 2018 RCT 10%-30% Yes Medium-high Amaral et al. 2021 RCT 27% Yes Medium-high Downloaded from https://academic.oup.com/wbro/article/40/2/211/7690325 by Laura Mowry user on 05 August 2025 Antelman et al. 2023 DD 20% Yes Medium Björkman Nyqvist and RCT 2% Not tested High Jayachandran (2017) Bjorvatn et al. 2022 RCT 4–8% 3–4 p.p. higher among control High Bos et al. 2024 RCT 3% Yes High Carneiro et al. 2021 RCT 23% Yes Medium-high Dinga 2019 RCT 4% Not tested High Doyle et al. 2018 RCT 6%∗ 2 p.p. higher among treatment High Fitzsimons et al. 2016 RCT 33–35% Yes Medium-high Jensen et al. 2021 RCT 10%∗∗ Not tested High Justino et al. 2023 RCT 10–18% Yes High Lassassi 2021 DD Not examined Medium Luoto et al. 2021 RCT 7% Yes High Ohrnberger et al. 2020 IV Higher∗∗∗ Depends on specification Medium Osaki et al. 2019 RCT 24–28% 4 p.p. higher among control High Özlüses and Çelebioglu 2014 DD Not analyzed Medium Rahman et al. 2008 RCT 11–12% 1 p.p. higher among control High Rosero and Oosterbeek 2011 RD Not relevant for study design Medium-high Wang et al. 2015 RCT 0% Yes High Wang and Lin 2019 DD 17% Not tested Medium Source: Authors’ analysis. Note: RCT = Randomized controlled trial. DD = Difference-in-differences. RD = Regression discontinuity. p.p. = per- centage points. When attrition is listed as a range, that indicates variation across time periods or across arms. Balance refers to whether rates of attrition are balanced across treatment groups. ∗ In Doyle et al., the 6 percent is attrition among men, the focus of this review. ∗∗ In Jensen et al., the 10 percent attrition is among caregivers. ∗∗∗ In Ohrnberger et al., the appendix reports that more than 4,000 individuals left the sample between the first two waves, more than 4,000 between the next two waves, and more than 2,000 between the subsequent two waves. The maximum individuals in the sample reported in any table is around 11,000. The World Bank Research Observer © The Author(s) 2024. Published by Oxford University Press on behalf of the International Bank for Reconstruction and Development / THE WORLD BANK. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted reuse, distribution, and reproduction in any medium, provided the original work is properly cited. https://doi.org/10.1093/wbro/lkae009