Today (July 3 2015) the ONS published a paper detailing the responses to the ONS consultation on the Approved Researcher Scheme. This document provides the basic evidence from which an Expert Steering Group were able to construct a detailed series of recommendations to the National Statistician to improve the existing Approved Researcher process. At the time of writing these recommendations have not been submitted, and I can’t pre-empt either the final set of recommendations or the National Statistician’s response to them. But I can say a little about how the UK Data Service has an impact in exercises like these, even though the Service did not formally make a response to the consultation.
Informing the consultation
In December 2014, I was approached to join a small steering group which had the primary aim of creating recommendations for changes to the Approved Researcher status. There were only three external representatives on this group, myself, and delegates of both the HSCIC and the Royal Statistical Society.
Before this group met in January, I approached all the key staff in the UK Data Service to give me a better understanding of some of the administrative challenges which the Service faces in providing access to data under the Approved Researcher scheme. This proved invaluable, and really helped inform the initial consultation document. Especially important, to my mind was our ability to examine the characteristics of researchers who had used the Secure Lab – we found that around 26% of researchers had been involved in more than one project, making a reasonably strong case for transferability of status across projects. We also discovered that less than 10% of users (over a five year period) had transferred institution, leading to a feeling that organisational indemnities were likely to last for a reasonable period.
Transferring Approved Researcher status?
At the first meeting of this group, a traditional open-ended consultation was proposed. However, after a fair amount of discussion, we came to the conclusion that a ‘straw man’ approach, where a series of proposals would be made, which may or may not have represented the desired outcomes of the group, was a better way to capture the feelings of the respondents to the survey, and make our job as an expert group easier. One of the key areas which I wanted to ensure was addressed implicitly in the consultation document related to the possibility of transferability of any Approved Researcher to any other legal gateway providing access to personal data.
The UK Data Service administers a number of processes which are more or less analogous to the Approved Researcher status, and it was important that these different administrative processes remain as harmonised as possible. The ONS were able to ask this explicitly. Question 17 on the AR criteria being promoted as the default access mechanism for all relevant bodies which give access to legally protected data, was agreed with by around 70% of the respondents, giving a clear steer to other relevant bodies.
This was one of the few recent consultations relating to data access which the Data Service hasn’t actually responded to formally, but we did publicise the consultation on our website, and we also made sure that all of the users of the Secure Lab were informed about it. We can’t say the extent to which we influenced people to participate, but the list of respondents suggests that we may have helped. This activity is one small example of how the UK Data Service works with others to make more data available to more researchers, while protecting the privacy of survey respondents. The results of the consultation are encouraging for this part of our mission, and we’ll wait to see to the outcome of the recommendations which this group has passed to the National Statistics Executive Group.