CSEA among infants and toddlers appears to be increasing. This data blueprint provides a path for measuring CSEA prevalence among this hidden population.

Data Blueprint for Measuring the Prevalence of CSEA in the Early Years

Children with shadow hands approaching

Current measures of CSEA prevalence miss a key population – children under five years old. All the more concerning when online data suggests that sexual abuse against infants and toddlers may be increasing.

This data blueprint aims to inform efforts to better measure the prevalence of CSEA within this hidden and vulnerable population. Childlight will use this as a roadmap for our future work so that the global evidence base around prevalence truly covers all of childhood.

Research findings

The research team found no age disaggregated studies on the prevalence of CSEA in the early years from a recently completed global systematic review.  This, however, does not mean this type of abuse is not happening. We conducted additional secondary analysis of big data from NGOs and police that track online child sexual abuse material globally and many URL links and dedicated websites to CSAM of infants and toddlers were discovered. The Internet Watch Foundation 2022 data shows that:

  • 1,001 URL sites lead to material depicting children ages 0-2 years old
  • 11, 351 URL sites lead to material depicting children 3-6 years old

We also know from data which focuses on online CSAM globally that:

  • This age group includes more severe abuse
  • There are concerns by those on the frontline response that this type of abuse is increasing
  • Child sexual abuse of infants and toddlers are present in more highly traded series of abuse images and videos (C3P, 2021; Interpol and ECPAT, 2018) 
Child with shadow hand

In many of these images and videos, the victims are still unidentified – meaning the abuse is still happening. Among a representative sample of unidentified victim images and videos from Interpol data, 60% of the unidentified victims were prepubescent (Interpol and ECPAT, 2018).

Ven diagram

Early years prevalence data blueprint

Data provides the foundation for consolidated action by policymakers, practitioners, and the public. The lack of robust prevalence estimates on CSEA in the early years hampers our collective response and ultimately means that abuse against society’s most vulnerable is allowed to continue.

The evidence shows us that survey estimates are not possible, and we must utilise both administrative and big online data to better understand the prevalence within this age group.

The biggest source of administrative data is police data. There will be significant overlaps between police data and online big data identified through NGOs. Where the location is identified within images and videos, this data is automatically handed over to local law enforcement. Some victims will be identifiable from policy records, and some won’t be. To measure prevalence, we will need to know the overlap between infants and toddlers that can be identified from both lists.

Police data also has overlap with other administrative data systems. These data sources include a country’s child protection data system (social welfare, child protection plans, etc), nursery and early childhood education referral data, child health and death data, as well as court data.

Multiple system estimation (MSE)

Based on this data blueprint, Childlight will utilise administrative datasets and a statistical method called Multiple Systems Estimation (MSE) to estimate the number of victims of sexual exploitation and abuse in the early years in the UK and overseas. This approach has been used previously to estimate the prevalence of modern slavery in the UK (Sharifi Far et al., 2021; Bird & King, 2018).

With MSE administrative data is collected in real time and does not rely on adult survivors' memories of childhood experiences. But there are challenges with MSE which include finding and accessing datasets that meet the requirements for analysis. Childlight is looking for administrative data owners who are interested in this approach and in estimating prevalence of CSEA in the early years to get in touch.