Data validity, reliability and confidentiality

To minimize bias and validate findings, we triangulated among data sources, always corroborating and comparing information from interviews with written sources, rather than relying predominantly on one or the other source of information. In reporting the interview data, we assigned each key informant a number, and listed their most prominent institutional affiliation type and country classification (see Table 2). To ensure historical accuracy and data validity, we incorporated feedback on a draft of this paper from four interviewees representing distinct perspectives. We also asked reviewers to identify any additional documents we should be reviewing that we had overlooked, and none indicated any. To enhance data reliability, interview audio was transcribed and the first author, who undertook all key informant interviews, conducted coding and analysis of the data collected. Both authors are outsiders to the children’s care community—social science researchers who initiated the research without any strong stance on any of the major global policy debates concerning children’s care.

Key informant number/organizational type

The study protocol underwent ethics review and received exemption by the Institutional Review Board of Johns Hopkins University. All interviews were recorded and transcribed with consent from participants. To ensure respondent privacy and confidentiality, we assigned each respondent a unique number (i.e., I(interview)1, I2, etc.) in coding the data and presenting the results in the manuscript so that they could not be linked to identifying information. We also ensured that all data collected (notes taken during the interview, interview audio, and transcribed interview) and documents outlining data collection approaches were stored in an encrypted, password-protected folder only accessible to authors of this manuscript.