Sometimes starting at step three or four makes step one easier when you circle back to it. When thinking about data sharing and system interoperability, there are many questions and issues to resolve. Certainly one of the primary issues to resolve is legal and another is around risk management. Both of which can be thorny and complex. And while sorting out the legal and risk issues is likely primary, step three might be the step in which you agree on what data will be shared.
Sometimes determining the data you want to share first enables you to more easily have the legal and risk conversations because you are talking in specifics. You will likely need to revisit the dataset after going through the legal and risk discussions, but by starting with a dataset you can ground the legal discussion in reality not hypotheticals.
The other advantage of talking about specifics as it allows you to have the conversation about purpose. So it’s hard to argue data sharing is a bad idea in general for humanitarian agencies, but it doesn’t actually explain the purpose.
If the purpose, as often assumed, is deduplication of datasets then perhaps we should start there as that is the ‘problem we are trying to solve’ rather than data sharing. And if deduplication is the true problem then perhaps other options, like data trusts or zero knowledge proofs, are better options. Data Trusts have been discussed here and the idea of a zero knowledge proof, in this case, is that one organisation could ‘ask questions’ of another organisation’s database. The answers would not contain the details of the data but would be yes/no.
All this to say that often our journey to answers of challenges is circular rather than linear and where we start is less important than we often think. It is the starting, the exploring, and the willingness to engage that is important.