Publications

This brief examines the potential ethical issues, including benefits and risks, associated with predictive analytics as they pertain to children. 
It is based on a more in-depth working paper, UNICEF Innocenti Working Paper 2021-08, which provides further detail, guidance, and tools.

This paper examines potential ethical issues, including benefits and risks, associated with predictive analytics as they pertain to children. It is designed to support readers in gaining an overview of the current state of the field, knowledge of real-world deployments of predictive analytics and ultimately, a deeper understanding of the opportunities and potential harms of deploying predictive analytics that directly or indirectly target children.

Children’s lives are increasingly mediated by digital technologies. Yet, when it comes to understanding the long-term effects of internet use and online experiences on their well-being, mental health or resilience, the best we can do is make an educated guess. Our need for this knowledge has become even more acute as internet use rises during COVID-19.

This report explores what has been learned from the latest research about children’s experiences and outcomes relating to the internet and digital technologies. It aims to inform policy-makers, educators, child-protection specialists, industry and parents on the best evidence, and it proposes a future research agenda.

This data-driven research brief explores three research questions. 1) How much do we know about children’s basic access to the internet across the globe? 2) Do children regularly use the internet to access health information? 3) Are children able to verify the truth of online information?

The brief analyzes survey data from the ITU World Telecommunications/ICT Indicators database, as well as household-survey data collected from approximately 22,000 children aged 12-16, generated by the collective work of the EU Kids Online and Global Kids Online research networks. It concludes with recommendations on how stakeholders can ensure that children’s health information needs are better supported during the COVID-19 pandemic and beyond

This paper provides a framework for analyzing constraints that apply specifically to women, which theory suggests may have negative impacts on child outcomes (as well as on women). We classify women’s constraints into four dimensions: (i) low influence on household decisions, (ii) restrictions on mobility, (iii) domestic physical and psychological abuse, and (iv) limited information access. Each of these constraints are in principle determined within households. We test the impact of women’s constraints on child outcomes using nationally representative household Demographic and Health Survey data from India, including 53,030 mothers and 113,708 children, collected in 2015-16. We examine outcomes including nutrition, health, education, water quality, and sanitation. In our primary specification, outcomes are measured as multidimensional deprivations incorporating indicators for each of these deficiencies, utilizing a version of UNICEF’s Multidimensional Overlapping Deprivation Analysis index. We identify causal impacts using a Lewbel specification and present an array of additional econometric strategies and robustness checks. We find that children of women who are subjected to domestic abuse, have low influence in decision making, and limited freedom of mobility are consistently more likely to be deprived, measured both multidimensionally and with separate indicators.