SIM 3 Wales: Maximising the use of existing administrative and survey data sets for Wales

Lead Research Organisation: Cardiff University
Department Name: Sch of Social Sciences

Abstract

The over-arching aim of the project is to review options for improving secure access to detailed administrative and survey data for Wales for research purposes. The first stage of the substantive research programme will be to describe the different surveys and administrative sources of data available within Wales building upon existing data audits and information held by the WG. The accessibility of these sources for the purposes of research will also be established in consultation with data custodians. How queries regarding the data are currently handled and how the governance arrangements surrounding applications to use the data are managed will be assessed. Where relevant, these arrangements will be compared to the arrangements that exist for comparable sources of data available elsewhere in the UK. An important aspect of the review will also be to identify sources of data that are potentially unique to Wales or where innovative links can be made to develop new sources. This will draw upon the findings of the other placement fellowships.
The second stage of the research programme will examine the potential benefits of improving access to the data sources identified in the audit. Existing users of this data (or similar data) will be consulted as to their perceptions regarding the potential usefulness of Welsh data. Those to be consulted will be identified from or through the membership of Thematic Reference Groups established by the Wales Institute for Social and Economic Research Data and Methods (WISERD). If the data audit identifies a source of data that is unique to Wales or potentially unique source of data that could be developed through data linking, researchers active in the field of study will be asked to consider whether their academic discipline could benefit from such data. The review will highlight areas where improved access to Welsh data or where opportunities for enhancing existing data may be expected to lead to significant benefits to the research community.
The third stage will identify the barriers to improving access to Welsh data. Barriers to access may simply be due to a lack of awareness among data custodians of the mechanisms for data access which are available or of the potential benefits that access to their data could yield. However, there may also be costs associated with providing access to data, such as the construction of analytical datasets, the development of supporting documentation or in supporting the complex governance arrangements surrounding access. These issues will be investigated through discussions with the owners and custodians of different data sets in order to provide a detailed understanding of the constraints faced.
The final stage of the research programme will identify the best ways to improve access to potentially disclosive data. The recommendations will seek to maximise the benefits that can be achieved from the greater utilisation of existing ESRC funded investments, such as the UKDA, SDS or ADLS. In other cases, the barriers to access may be greater. Here, recommendations will reflect on the lessons learnt from initiatives developed elsewhere in the UK to provide research access to highly sensitive data. Experts from these bodies will be identified and invited to form the basis of a consultative group whose knowledge can be drawn upon in the development of these recommendations.
The research programme will culminate in a report that provides a detailed review of Welsh data sources. Specific action plans that outline the steps that would need to be taken to improve access to different data sets will also be provided. In each case, key contacts and bodies that can support data custodians in implementing these measures will be identified. In addition to data specific recommendations, protocols will also be developed in order to provide a suggested uniform approach to data access policy that can be implemented across WG and related bodies.

Planned Impact

In the short term (12 months), the main beneficiaries of the scheme will be the WG who will be provided with a report that provides well considered and detailed recommendations regarding how the benefits from existing investments in WG funded data sources can be maximised. It is envisaged that the report will form the basis of an integrated and consistent approach to data access both across WG and within other bodies that are linked to the WG. The fellowship will help place the WG at the forefront of developments regarding the implementation of measures that seek to maximise the benefits that can be gained from existing data resources through improvements in access and opportunities for data linkage. This programme will therefore also be of interest to other government departments in informing their data access strategies.
It is acknowledge that the main benefits of the research programme are dependent upon whether or not the WG accepts the recommendations outlined in the final report. The active engagement of the academic community, data custodians and experts in the fields of data access and research support should ensure that the recommendations arising from the research programme are 1) sensitive to constraints of data custodians, 2) proportionate to the potential benefits that would arise from enhancements to data or data access and 3) practical. This approach should ensure that the recommendations are well considered and have the greatest chance of being implemented.
Irrespective of this, the programme of research will be of direct benefit to policy makers across Wales. The data audit will raise awareness of existing data sources in Wales. If agreed with the WG, this could be published as a 'stand alone' reference volume for wider research community. Policy makers within WG, bodies affiliated to WG, local government or other public bodies will have greater awareness of data sources and could commission research to inform their decisions across a number of policy areas, reinforcing the importance of realising the maximum value out of existing data sources before commissioning new data collection exercises. This may be of particular importance to organisations where analytical resources are relatively limited. Such bodies will particularly benefit from guidance in relation to data collection and access that will arise out of the fellowship and the increased levels of evidence that analysis of the data that they collect will provide.
Discussions with data custodians and policy makers within WG and other bodies will raise awareness and engagement of such groups with UK wide developments on data access. Discussions with ONS and experts from bodies that support and promote data access will also contribute to the increased consideration of issues specific to Wales in relation to data access arrangements. The engagement of both the academic community and public bodies in the research programme will help to further develop relations between both groups.
Subject to the recommendations taken on board, in the longer terms (12 months+) a number of public bodies in Wales will begin to benefit from the programme. The utilisation of well established mechanisms for the release and access of data will also improve the security with which data about Wales is accessed and monitoring of how that data is used. In the longer term, the research conducted on these sources will provide WG with greater understanding of the economic and social circumstances of people living in Wales beyond that which is possible for the WG to achieve 'in-house' with limited resources. Several data sets will have been developed directly in response to monitoring or evaluation requirements, such as surveys of participants or businesses supported by projects funded by European Structural Funds. As such, the analysis of these sources will be of direct interest to policy makers within WG, the UK and in some cases Europe.

Publications

10 25 50
 
Description This Report is delivered as part of the WG Programme to Maximise the Use of Existing Data and aims to examine issues surrounding the feasibility of establishing a Wales Longitudinal Study. The existing Census-based UK Longitudinal Studies are the Office for National Statistics England and Wales LS (1% sample of the population), Scottish LS (5% sample) and Northern Ireland LS (28% sample). Whilst the design, content and coverage of these studies varies, it is the inclusion of Census data that is crucial to their importance as a research resource. Unlike other devolved administrations in the UK, the Welsh Government does not have legal authority for conducting the Census in Wales, with ONS being responsible for conducting the England and Wales Census. This poses a significant barrier to the establishment of Wales Longitudinal Study similar to those established in other devolved nations. The absence of such a resource limits the ability to conduct research about the population of Wales in areas where law making powers have been devolved to the Welsh Government.


The two primary aims of this project have been to:
• Identify and evaluate the options for creating a 100% 'population spine' to facilitate analysis for detailed population sub-groups and to act as a basis for linking administrative and survey data sets for Wales, with particular attention being given to examining issues surrounding the inclusion of Census data in to such a study.
• Examine the extent to which the use of 'richer' social survey data sets can be maximised through using statistical matching or regression-based imputation methods to 'fuse' retrospective anonymised survey data to the 'population spine' in order to create a simulated population of Wales.
The project included a literature review, discussions with key individuals in Government Statistical Agencies and Data Access Infrastructures and exploratory analysis of relevant administrative and survey data sets. This project is closely related to the ONS Beyond 2011 (B2011) Census Replacement Programme and the work of the UK Administrative Data Taskforce (ADT). The recommendations of the B2011 Programme could have a significant impact on the future landscape of Census longitudinal studies and the data that are available.
Exploitation Route The responsibility of ONS for conducting the Census in Wales and their responsibilities to Census respondents as custodians of Welsh Census data has practical implications in terms of where and how a WLS could be constructed. The desired WLS sample size of 100% would also have implications in terms of public acceptability and maintaining the confidentiality of respondents. Whilst it would not be feasible to create a 100% Census-based WLS as currently constituted, several options that warrant further examination remain:
• In the absence of a 100% WLS, WG could work with ONS to explore whether mechanisms exist to make better use of the England and Wales LS, including as a spine to link administrative data. Whilst the sample size of the ONS LS may not support Wales-level analysis for detailed population sub-groups, projects that focus on linking administrative data for broader groups could be useful in highlighting the research potential of a Census based WLS.
• WG should consider whether it would be desirable to commission the ONS and the Health and Social Care Information Centre to produce a Census based WLS of sufficient size to support Wales-level analysis.
• WG should consider the recommendations made by B2011 and the UK ADT in terms of maximising the use of administrative data as a population spine to inform the development of an alternative or complementary non-Census based administrative database.
• Administrative databases tend not to include detailed socio-economic data such household structure or social class. These problems present a significant practical and methodological challenge to the application of statistical matching or imputation techniques to a WLS where a population spine was based on administrative data.
• There is currently uncertainty surrounding the recommendations that will be made by the Beyond 2011 Programme regarding the future design of the Census. However to be able to deliver a Welsh Longitudinal Study based on the Census, the Welsh Government would need to seek whether it can access Census data under future arrangements or consider whether they would wish to take responsibility for the Census in Wales.
Sectors Communities and Social Services/Policy,Education,Government, Democracy and Justice

URL http://wales.gov.uk/statistics-and-research/examining-feasibility-creating-wales-longitudinal-study/?lang=en
 
Description Informed discussions within WG regarding the Beyond 2011 Census Programme; the establishment of a Wales Longitudinal Study and possibilities within WG Data Maximisation Programme. The findings of the report also provided valuable material in terms of the development of a successful bid for the formation of a Wales Administrative Data Research Centre by Swansea and Cardiff Universities.
First Year Of Impact 2013
Sector Communities and Social Services/Policy,Education,Government, Democracy and Justice
Impact Types Policy & public services

 
Description Rhys Davies (WP3.3) 25.11.16 - Making Sense of Official Estimates of Trade Union Membership, Labour Force Survey Users Conference, UK Data Service, Manchester University 
Form Of Engagement Activity A talk or presentation
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Other audiences
Results and Impact Conference paper delivered at the UK Dataservices' Labour Force Survey Users Conference on 25 November 2016. The user conference brings data producers and data users together to share updates on the development of the surveys and to showcase research that is being carried out using the data. The programme will contain a mixture of presentations from the Office for National Statistics (ONS) and researchers who use data from the Labour Force Survey (LFS) and Annual Population Survey (APS). https://www.ukdataservice.ac.uk/media/604639/davies.pdf
Year(s) Of Engagement Activity 2016
URL https://www.ukdataservice.ac.uk/news-and-events/eventsitem/?id=4729