Empirical evidence from across the globe demonstrates that the foundations for human development are laid down in early childhood, the first thousand days after conception being the most crucial for determining later outcomes. The circumstances an individual experiences during this period make a huge difference to their later development. This raises significant concerns about the countless children throughout the world who are born into poverty, despite the strong performance of many economies in recent years. More worrying still, deprivations frequently happen together, with cumulative effect. Thus, poor children commonly have limited or no access to services needed for sound health and development, including potable water, sanitation, quality health care and education. They are also regularly exposed to other less often documented risks, such as violence, overcrowding, noise pollution, toxins and lead. In many contexts, poverty and associated adversities in the first phase of life account for the early emergence of substantial developmental disparities in learning and other outcomes children in different wealth groups.

Early risk exposure can have profound and lasting consequences, especially given that different developmental processes in humans are interconnected. Depending on the age of the child, one form of poverty-related deprivation may impact multiple developmental domains and functions, resulting in cumulative consequences (a developmental cascade) for development over time and generation. For example, chronic malnutrition in the first two years is likely to negatively impact the development of the child’s neurological architecture and functioning, which in turn limits learning potential, compromises education outcomes and ultimately reduces job prospects, and holds the next generation in poverty. Possibly most worrying of all, some risks experienced in early life, particularly nutritional deprivation—manifested as stunting— are held to be irreversible once the critical thousand days’ window has passed.

In light of this evidence, the current academic and policy attention to the early years seems fully justified. Nevertheless, this begs the question as to whether early childhood is the only critical period in a child’s life, or whether—and to what degree—later remediation of prior risk exposure is possible. These questions are vitally important because currently we do not seem able to prevent large numbers of infants from being exposed to highly detrimental circumstances; focusing exclusively on the first thousand days of life offers no hope to children who are born into and raised in detrimental circumstances. Besides, even though early childhood shocks generally have persistent effects this does not mean that they have to; indeed a number of studies now find a substantial degree of ‘catch-up growth’ among children who were initially stunted. While we still do not fully understand all aspects of catch-up growth – in particular, causal studies are few – the existing body of evidence does suggest that some recovery is possible. This can even be linked directly to policy measures: for example, Singh Park and Dercon (2014) find that the Midday Meals Scheme in India compensated at the age of 5 years for nutritional deficits in early childhood associated with a very severe drought. Understanding which factors contribute to recovery, when in a child’s life recovery is possible and why some children are more resilient than others is vital for designing remedial interventions.

Similarly, even though shocks during the early years impact subsequent outcomes, this does not account for the lion’s share of variation between children in later life. For example, while differences in learning levels across four countries (Ethiopia, India, Peru and Vietnam) emerged early and were evident even before school-entry age, they grew remarkably after children entered school and this trend seems to be almost entirely a product of how much children learnt per year of schooling. Even with initial disparities in learning at age 5, simply equalizing how much children learned per year of schooling would entirely close the gaps that appeared at the age of 8 years between children in Vietnam and in Peru: in 2012, Vietnam was among the highest scoring countries in the PISA assessment while Peru was the lowest. This does not mean that learning disparities emerging in the early years do not matter – they clearly do – but it does indicate that, depending on the domain, there may well exist critical periods for intervention later on in childhood.

Although this evidence offers crucial insights that are broadly applicable to policy globally, it is also clear that contexts can matter in powerful ways that are vital in determining who interventions should be aimed at. The same economic shock can have a very different effect across types of households, and within them, across individuals. Indeed, the risks facing children may often be specific to contexts, domain and ages. For example, studies in different settings have now documented that outcomes, both immediate and long-term, of negative income shocks in early childhood are more sensitive for females than they are for males. In India, the effect of early childhood income shocks on mortality seems only evident for girls, as do the effects on eventual adult height, health and assets in Indonesia, suggesting that compensatory behaviour or the lack of, by households has a significant role to play later. But this does not imply that gender gaps are universally present or always favour boys. In a recent paper, we document that this is in fact not the case for a wide suite of indicators across countries.

Children feature significantly in current discussions around the Post -2015 Sustainable Development Goals. This is an encouraging sign. However, if the resulting interventions are to be effective we need to expand the evidence base concerning the factors that promote and impede children’s development across developmental domains and functions, contexts and age groups.  A more methodological point, often overlooked, relates to the types of data we choose to invest in to learn about child development. Specifically, many of the crucial insights on child development – for example the complementarities between systems or on resilience and recovery – are based on a relatively small number of cohort studies (or panel data more generally), not just in low and middle-income countries but also in high-income countries such as the US (see, for example, the seminal work of Jim Heckman, Flavio Cunha and co-authors). These studies make at least three distinct contributions: they provide a clearer understanding of the dynamics of child development, both across and within domains; they further motivate the development of interventions which can be tested experimentally; and pragmatically, they are often the best (sometimes the only) way of evaluating large changes that happen unplanned. That only a few countries have made these long-term investments in data remains an important cause for concern, especially given that policies and interventions around child development in many contexts are severely hampered by assumption, conjecture and knowledge gaps.

This article was first published by the World Bank’s Let’s Talk Development blog. Publication does not imply endorsement of views by the World Economic Forum.

To keep up with Forum:Agenda subscribe to our weekly newsletter.

Author: Jo Boyden is the Professor of International Development and Director, Young Lives, University of Oxford. Abhijeet Singh is a D.Phil. (Ph.D.) student in Economics, and a Research Officer at the Department of International Development (QEH), at the University of Oxford.

Image: Kelvin Villaroel holds up his son Isaac for a photo in front of a cherry tree in full blossom in Central Park in New York April 20, 2014. REUTERS.