Research

Coverage: VoxDev, Ideas For India, Discover Economics.

Abstract: This paper explores how parents in Rajasthan, India decide how long to keep a daughter in school and when, and to who, to get her married. We develop a dynamic model of parental decision making in which parents take schooling and marriage decisions under uncertainty about the future marriage offers they will receive. Parents' choices are thus partially driven by their beliefs about the likelihood of receiving high-quality marriage offers in the future and how this likelihood depends on a daughter's age and education. We use this model to create a novel hypothetical-choice tool. By varying the degree of certainty over future marriage offers described in hypothetical scenarios, our tool enables us to identify both preferences and probabilistic beliefs without directly eliciting probabilities. We find that parents believe there to be a large marriage-market return to girls' education and that this drives much of the investment that parents make in their daughters' education. And while parents would prefer, other things being equal, to delay a daughters' marriage until at least age 18, a belief that marriage-market prospects begin to deteriorate with age once a daughter is out of school creates an incentive to parents to accept early marriage offers. 

Abstract:  The social environment is key to sustaining gender inequalities but many policies and programs target only women and do not involve the wider community. Can such approaches work or, by pushing women to break accepted norms, do they expose women to stress and backlash? What are the impacts of engaging the wider community? We use a 3-armed RCT covering 5000 adolescent girls across 125 communities in rural Rajasthan to explore these questions. We assess the impacts of weekly girl groups that only worked with adolescent girls and the impacts of additionally engaging the wider community.  Both models led to a reduction in school dropout and early marriage. However, targeting adolescent girls without involving the broader community led to an increase in girls adopting a ruminative thinking style and no improvements in depression and anxiety. By contrast, when the wider community was engaged, girls' symptoms of depression and anxiety fell by 0.26 SD and 0.33 SD respectively and negative impacts on rumination were reversed. We show evidence that such improvements in mental health may have resulted from the community engagement changing prevailing attitudes, norms and the sanctions girls perceived they would face for breaking norms.

Coverage: The 74

Abstract: Global access to preschool has increased dramatically, yet preschool quality is often poor and evidence on how to improve it is scarce. We worked with the government of Colombia to implement a large-scale randomized controlled trial evaluating two interventions targeting the quality of public preschools in Colombia. The first, which was designed by the government and rolled out nationwide, provided preschools with significant extra funding, mainly earmarked for hiring teaching assistants (TAs). The second additionally offered professional development training for existing teachers, delivered using a novel low-cost video-conferencing approach. We find that, despite increasing per-child expenditure by around a third, the first intervention did not improve child development and led to a reduction in the time that teachers spent in the classroom, including on learning activities. In contrast, the second intervention led to significant improvements in children’s cognitive development, especially those from more disadvantaged backgrounds, at little extra cost. The addition of the professional development training offset the adverse effects of TA provision on the time teachers spent on learning activities in the classroom and improved the quality of teaching. When we interpret our results through the lens of a model of teacher behavior, two insights arise. First, income effects and a perception that TA time was a good substitute for their own may have led teachers to endogenously scale back their efforts in the classroom in response to the provision of new resources. Second, the training prompted teachers to increase their perception of the usefulness of learning activities for child development and their perception that they had a comparative advantage in these learning activities relative to the TAs.

Abstract: We study the effects of incentivizing individuals to use healthcare in the presence of congestion externalities. Our theoretical model highlights interesting interactions between the size and shape of the congestion externality, and what drives selection into institutional delivery. We focus on a conditional cash transfer program (JSY) in India that paid women to give birth in medical facilities. In areas with below-median health-system capacity, JSY increased perinatal mortality. We provide evidence that a congestion externality was a key driver of these harmful effects. Moreover, JSY reduced childhood vaccination rates suggesting a diversion of resources away from routine services. 

Coverage: VoxDev

Abstract: Social connections are fundamental to human wellbeing.  This paper examines the social networks of young married women in rural Odisha, India.. This is a group, for whom highly-gendered norms around marriage, mobility, and work are likely to shape opportunities to form and maintain meaningful ties with other women. We track the social networks of 2,170 mothers over four years, and find a high degree of isolation. Wealthier women and women more-advantaged castes have smaller social networks than their less-advantaged peers. These gradients are primarily driven by the fact that more-advantaged women are less likely to know other women within their same socioeconomic group than are less-advantaged women are. There exists strong homophily by socioeconomic status that is symmetric across socioeconomic groups. Mediation analysis shows that SES differences in social isolation are strongly associated to caste, ownership of toilets and distance. Further research should investigate the formation and role of female networks.

Abstract: We explore the educational experiences of Indian children during the pandemic, using time-use and household expenditure data from a representative panel of over 230,000 households. We find that both 12-18-year old’s average learning time and their average households’ expenditure on education more than halved following the March 2020 school closures. Both had barely recovered by the end of 2021, and appear remarkably unaffected by phased school reopenings. Interpreting the changed patterns of educational investments through a simple model of skill formation suggests skill inequalities between cohorts may increase, while implications for within-cohort inequalities are ambiguous. Children from households who experienced more-severe economic shocks during the pandemic saw larger losses in inputs although heterogeneity by socio-economic characteristics is more mixed. Overall, differences in losses across subgroups are dwarfed the average losses: every subgroup we analyze experienced average falls in learning time and educational expenditure, respectively, of at least 42% and 59%.

(Previous Briefing Note version: How are mothers and fathers balancing work and family under lockdown?) 

Coverage: Economist, BBC, The Guardian, Sky News, iNews, The Telegraph, The Independent, ITV News, Today Programme (BBC Radio 4), BBC Breakfast, BBC Television News. 

Abstract: COVID-19 has uprooted many aspects of parents’ daily routines, from their jobs to their childcare arrangements. In this paper, we provide a novel description of how parents in England living in two-parent opposite-gender families are spending their time under lockdown. We find that mothers’ paid work has taken a larger hit than that of fathers’, on both the extensive and intensive margins. We find that mothers are spending substantially longer in childcare and housework than their partners and that they are spending a larger fraction of their paid work hours having to juggle work and childcare.  Gender differences in the allocation of domestic work cannot be straightforwardly explained by gender differences in employment rates or earnings. Very large gender asymmetries emerge when one partner has stopped working for pay during the crisis: mothers who have stopped working for pay do far more domestic work than fathers in the equivalent situation do. 

Abstract: We explore gender inequalities in paid and unpaid work with a focus on the UK. The average working-age woman in the UK earned 40% less than her male counterpart in 2019; Sizeable gender gaps in participation, hours worked and hourly wages all contribute to this gap. We explore how these patterns have changed over the past 25 years and conclude that after accounting for women’s rising education, progress has been modest. At the same time, women do far more unpaid domestic work than men.  We show how inequalities evolve around parenthood and highlight how the division of labor between parents appears remarkably unrelated to relative earnings potential. Gender norms, preferences and beliefs appear crucial. We discuss consequences of our findings for material inequality and for the economy at large. We discuss the likely impacts of various current and potential policies, including parental leave, childcare and the tax and benefit system.

Coverage: The Guardian, BBC, TES, The Independent, Metro, Bloomberg, Sky News, The Telegraph, , Today Programme (BBC Radio 4), BBC Television News.

Abstract: This paper combines novel data on the time use, home-learning practices and economic circumstances of families with children during the COVID-19 lockdown with pre-lockdown data from the UK Time Use Survey to characterise the time use of children and how it changed during lockdown, and to gauge the extent to which changes in time use and learning practices during this period are likely to reinforce the already large gaps in educational attainment between children from poorer and better-off families. We find considerable heterogeneity in children's learning experiences – amount of time spent learning, activities undertaken during this time and availability of resources to support learning. Concerningly, but perhaps unsurprisingly, this heterogeneity is strongly associated with family income and in some instances more so than before lockdown. Furthermore, our analysis suggests that any impacts of inequalities in time spent learning between poorer and richer children are likely to be compounded by inequalities not only in learning resources available at home, but also in those provided by schools.

Abstract: 

Background: An estimated 63.4 million Indian children under 5 years are at risk of poor development. Home visits that use a structured curriculum to help caregivers enhance the quality of the home stimulation environment improve developmental outcomes. However, achieving effectiveness in poor urban contexts through scalable models remains challenging.  

Methods: Using a cluster randomised controlled trial, we evaluated a psychosocial stimulation intervention, comprising weekly home visits for 18 months, in urban slums of Cuttack, Odisha, India. The intervention is complementary to existing early childhood services in India and was run and managed through a local branch of a national NGO. The study ran from August 2013 to July 2015. We enrolled 421 children aged 10–20 months from 54 slums. Slums were randomised to intervention or control. Primary outcomes were children's cognitive, receptive language, expressive language and fine motor development assessed using the Bayley-III. Prespecified intent-to-treat analysis investigated impacts and heterogeneity by gender. Trial registrations: ISRCTN89476603, AEARCTR-0000169.  

Results: Endline data for 378 (89.8%) children were analysed. Attrition was balanced between groups. We found improvements of 0.349 of a standard deviation (SD; p = .005, stepdown p = .017) to cognition while impacts on receptive language, expressive language and fine motor development were, respectively, 0.224 SD (p = .099, stepdown p = .184), 0.192 SD (p = .085, stepdown p = .184) and 0.111 (p = .385, stepdown p = .385). A child development factor improved by 0.301 SD (p = .032). Benefits were larger for boys. The quality of the home stimulation environment also improved.  

Conclusions: This study shows that a potentially scalable home-visiting intervention is effective in poor urban areas.

Abstract: 

Background:  Poor early childhood development (ECD) in low- and middle-income countries is a major concern. There are calls to universalise access to ECD interventions through integrating them into existing government services but little evidence on the medium- or long-term effects of such scalable models. We previously showed that a psychosocial stimulation (PS) intervention integrated into a cash transfer programme improved Colombian children’s cognition, receptive language, and home stimulation. In this follow-up study, we assessed the medium-term impacts of the intervention, 2 years after it ended, on children’s cognition, language, school readiness, executive function, and behaviour.

Methods and findings: Study participants were 1,419 children aged 12–24 months at baseline from beneficiary households of the cash transfer programme, living in 96 Colombian towns. The original cluster randomised controlled trial (2009–2011) randomly allocated the towns to control (N = 24, n = 349), PS (N = 24, n = 357), multiple micronutrient (MN) supplementation (N = 24, n = 354), and combined PS and MN (N = 24, n = 359). Interventions lasted 18 months. In this study (26 September 2013 to 11 January 2014), we assessed impacts on cognition, language, school readiness, executive function, and behaviour 2 years after intervention, at ages 4.5–5.5 years. Testers, but not participants, were blinded to treatment allocation. Analysis was on an intent-to-treat basis. We reassessed 88.5% of the children in the original study (n = 1,256). Factor analysis of test scores yielded 2 factors: cognitive (cognition, language, school readiness, executive function) and behavioural. We found no effect of the interventions after 2 years on the cognitive factor (PS: −0.031 SD, 95% CI −0.229–0.167; MN: −0.042 SD, 95% CI −0.249–0.164; PS and MN: −0.111 SD, 95% CI −0.311–0.089), the behavioural factor (PS: 0.013 SD, 95% CI −0.172–0.198; MN: 0.071 SD, 95% CI −0.115–0.258; PS and MN: 0.062 SD, 95% CI −0.115–0.239), or home stimulation. Study limitations include that behavioural development was measured through maternal report and that very small effects may have been missed, despite the large sample size.

Conclusions: We found no evidence that a scalable PS intervention benefited children’s development 2 years after it ended. It is possible that the initial effects on child development were too small to be sustained or that the lack of continued impact on home stimulation contributed to fade out. Both are likely related to compromises in implementation when going to scale and suggest one should not extrapolate from medium-term effects of small efficacy trials to scalable interventions. Understanding the salient differences between small efficacy trials and scaled-up versions will be key to making ECD interventions effective tools for policymakers.

Abstract: In Colombia’s bottom socio-economic strata, 46.6% of children under two are anaemic. A prevalence of above 20% falls within the WHO guidelines for daily supplementation with multiple micronutrient powder (MNP). To evaluate the effect of daily MNP supplementation on anaemia amongst Colombian children aged 12–24 months we ran a cluster RCT (n=1440). In previous work, we found the intervention had no impact on haemoglobin or anaemia in this population. In this current paper, we investigate this null result and find it cannot be explained by an underpowered study design, inaccurate measurements, low adoption of and compliance with the intervention, or crowding out through dietary substitution. We conclude that our intervention was ineffective at reducing rates of childhood anaemia because MNP itself was inefficacious in our population, rather than poor implementation of or adherence to the planned intervention. Further analysis of our data and secondary data suggests that the evolution with age of childhood anaemia in Colombia, and its causes, appear different from those in settings where MNP has been effective. Firstly, rates of anaemia peak at much earlier ages and then fall rapidly. Secondly, anaemia that remains after the first year of life is relatively, and increasingly as children get older, unrelated to iron deficiency. We suggest that factors during gestation, birth, breastfeeding and early weaning may be important in explaining very high rates of anaemia in early infancy. However, the adverse effects of these factors appear to be largely mitigated by the introduction of solid foods that often include meat. This renders population wide MNP supplementation, provided after a diet of solid foods has become established, an ineffective instrument with which to target Colombia’s childhood anaemia problem.