The Royal Statistical Society has produced the following response to the discussions and proposals on data sharing for research and statistics:
Our National Statistics Advisory Group has welcomed close involvement in the open policy process facilitated by the Cabinet Office with Involve. We have seen sensible discussion of options for research and statistics with a wider range of government and non-government stakeholders than might otherwise be the case. Omitting other strands of work on “tailored public services” and “fraud, error and debt”, in which we have not participated, we have the following comments.
Documents as currently drafted recognise the aspiration to develop ‘broad powers’ in legislation that it is thought would encourage more data to be linked for research and statistics. However they also recognise that attempts to introduce broad powers in legislation can carry greater risks and can lead single reforms that would otherwise be acceptable to fail. For example public support for data to be linked together for one purpose may be countered by opposition to linkage for other purposes. Support improves however when a clear case for public benefit is made. For research and statistics the limits of activities enabled from the proposed changes can be shown and we feel that starting from this, the case for public support could be well made.
This process has also built some understanding of the technical processes being proposed to de- identify data for research, and the models being favoured. Models should be subject to review in practice but building understanding in this way is valuable so that they are taken up and tested.
We also note some of the limits of the scope of this policy-making exercise, firstly with regard to health and care data and secondly with regard to the private sector. The paper on de-identified data states that the Department of Health “does not believe it is appropriate for another or alternative method of disclosure [of health and care data] to be made available”. Linking health data with data from other departments is known to address research questions in the public interest, so potential barriers to such questions being posed warrants discussion elsewhere. Sustainable legislation would also need to address the scope of the sector’s activities. We welcome the statement that legislation emerging from this particular policy process will not speak against private bodies or persons being researchers. Work elsewhere will be valuable to address some of the issues raised by private sector involvement.
Finally, we would share the view that policy-making and research is no substitute for public consultation on any legislation that follows. This is expected to take place separately and in full.
The full letter can be accessed via: RSS_Letter_to_CO_data_sharing_team_Aug2014-2