Assessing the impact of benefit sanctions on health

Lead Research Organisation: University of Glasgow
Department Name: School of Social & Political Sciences

Abstract

The aim of this research is to examine whether benefit sanctions lead to claimants having worse physical or mental health, or making greater use of health services. It also seeks to add to our knowledge on whether sanctions encourage claimants to return to employment more quickly. The proposal is innovative in using a database of individuals' benefits, employment and health histories constructed from administrative records which have not previously been linked.

Background
People who claim some benefits, especially Jobseeker's Allowance (JSA) and Employment & Support Allowance claimants in the Work-Related Activity Group (ESA WRAG), have conditions placed on them. With JSA, for example, claimants must sign on at the Jobcentre Plus office regularly and take certain steps to find work. If they fail to meet these conditions, they can be sanctioned, i.e. their benefits are stopped for a period. There have always been conditions on unemployment benefits but they have increased in recent years, with the number of sanctions and maximum length rising. The introduction of Universal Credit has seen conditions extended to those in work as well as higher sanction rates for early cases.
Sanctions are meant to ensure that people return to work as soon as they can. This is good for public finances as it keeps claims down, good for the economy as it keeps labour supply up and, arguably at least, good for claimants since unemployment tends to be bad for our health. However, critics argue that sanctions may have unintended side-effects: harming claimant health, increasing homelessness, or putting stress on families which affects child welfare. Health effects may arise from restricted spending on food or heating but also from the psychological stresses of trying to cope without income. These unintended impacts could lead to greater expenditure on public services, off-setting savings from reduced benefit claims. There has been very little research on these unintended impacts associated with sanctions, and almost none which can provide convincing evidence that sanctions cause harm.

Proposed research
We will construct a database of individuals' benefit, work and health histories by combining information held in the administrative records of the Department for Work and Pensions (DWP) and NHS Scotland. The data are sensitive, personal information so we need special permissions to access them which we are in the final stages of securing. The database will cover people who claimed JSA or ESA WRAG in Scotland for 2010-17. We focus on Scotland because the health data are currently only available here. From DWP, we get records of claim spells and whether people were sanctioned. From NHS Scotland, we get information on: the prescribing of medicines which can indicate various physical and mental health conditions; use of Accident & Emergency or unplanned hospital admissions which can indicate incidents arising from risky behaviours such as alcohol abuse or the worsening of chronic conditions due to poor diet, stress, etc.; and mortality data, including suicides. We will also examine missed appointments since these are a significant concern for the health services.
From Oct 2013 on, DWP can provide data on employment spells so we can look at the labour market impact of sanctions in this period, and hence whether any health effects depend on whether people find paid work or not.
The main challenge for our analysis is to show whether sanctions cause a change in health. To do this, we have to rule out the possibility that there are other factors which mean that the people who are more likely to get sanctioned also tend to have worse health; examples might include a drug or alcohol problem, or a domestic crisis such as a bereavement. There are various techniques which let us do this.

The results of the research will feed into academic debates about the recent changes to welfare but are also intended to have an influence policy in this area.

Planned Impact

This research has clear potential for economic and societal impacts, through its influence on welfare policy and debates, but also by demonstrating the value of linked administrative data for research.

The most direct beneficiaries of our research are current and future welfare benefit claimants. Evidence that the current sanctions regime has harmful impacts on physical and mental health would provide a powerful impetus for reforms to reduce those harms. The research will also provide new knowledge on the effectiveness of sanctions in relation to employment outcomes which could improve the ability of the system to support returns to work. A range of third sector organisations have campaigned on what they argue are the negative impacts of current sanctions policies for social welfare, including health. To date, they have drawn largely on qualitative research and case studies of their service users. Valuable as this is in identifying areas of concern, such research cannot provide convincing evidence of the causal impacts of sanctions. This work would provide important new information which could be used by these groups to inform public debates.

It should be stressed that the research may support or challenge the claims made by campaigning groups. There is valuable impact in either case: in providing evidence for change or a justification for current policies.

Evidence of impacts on health and health service use from sanctions would be of wider value to the general public as funders and users of these services. The research will examine whether sanctions put additional pressure on A&E services as well as hospital admissions, and whether they lead to more missed out-patient appointments. Indirectly, it looks at pressure on GP services, as increased prescribing rates indicate increased attendance at GP surgeries. All have been seen as being under great pressure in recent years. The health bodies responsible for providing these services will be interested in evidence of the impacts of welfare reforms on their activities, and likely to use these in future discussions with DWP.

Improving employment outcomes could produce fiscal gains as well as economic benefits from increased labour supply. There will be particular interest here from the devolved administrations such as the Scottish Government and from English local authorities whose funding is increasingly linked to the economic performance of their regions.

The identification of unintended impacts on health would bolster the case for using a similar approach to look at other potential side effects discussed in the academic literature and in public debates, notably on criminal justice and homelessness systems, and on family and child welfare. More fundamentally, it would add powerfully to arguments that welfare policy needs to consider potential unintended consequences and review evidence on likely effectiveness from the outset.

Lastly, by demonstrating what can be achieved through the analysis of linked administrative data, the research could bolster support for this kind of work and hence widen opportunities for similar work in the future, offering diverse potential benefits for policy making and public debate. At root, the ability to exploit linked administrative data relies on public support for data sharing. A clear case study of the benefits which can be achieved from data linkage research has the potential to boost such support and to reinforce to political leaders and public bodies the value to be gained from supporting this work. This project will be particularly valuable in this respect because of the novelty of the data linkages proposed and the high profile of the issues being examined.
 
Description Work on this project has been delayed due to factors wholly outwith our control. We have been granted an 18-month no-cost extension. Pending delivery of data, we have progressed work in three directions.
First, as a preparatory analysis, we have analysed the variation in benefit sanctions rates across Jobcentre Plus offices. Strong, persistent variation might be a useful basis for an instrumental variable for the subsequent causal modelling. However, our findings for Scotland provide little evidence of consistent differences in benefit sanctions rates between Jobcentre Plus offices over time. The findings from this research has been used to refine the design of our research project based on cross-sectoral individual administrative data. The findings were also used to inform an evaluation study of the labour market impacts of benefit sanctions which is being developed by colleagues from the Department for Work and Pensions.
Second, we are making a systematic assessment of the existing quantitative literature to understand the current state of the evidence on the impacts of benefit sanctions on labour market and wider outcomes. Our initial findings from a scoping review of international quantitative studies shows that most studies on the wider impacts of benefit sanctions used non-experimental designs and provided little support for the identification of any causal effects. These findings have sparked a fruitful discussion at a European Network for Social Policy Analysis (ESPAnet) conference and attracted the interest from leading analysts from the Department for Work and Pensions.
Third, we conducted some exploratory analysis of administrative data linked to social survey data in order to understand the quality of administrative data which will be delivered under our data sharing request. This resulted in the journal article (Pattaro et al 2019). We showed that longitudinal administrative data provided a picture of labour market careers which was consistent with survey recall questions, and that they were at least as good as those questions in predicting current poverty risks. The findings provide reassurance that administrative data from the DWP will prove a useful basis for our main project.
Exploitation Route The results of the scoping review will potentially have widespread value for those developing welfare conditionality policies in the UK and elsewhere, as well as those interested in this policy area from the public and third sectors. The latter include people working or interested in closely related areas where the impacts of welfare conditionality may be felt. That includes those in health, education, housing, social service and criminal justice. We will seek to disseminate results widely at the appropriate time.
The results of the analysis of linked admin and survey data provide general support for continued efforts to widen access to administrative data, under the ADR-UK umbrella.
Sectors Communities and Social Services/Policy,Healthcare,Other

URL https://www.scadr.ac.uk/our-research/work-and-social-security/assessing-impact-benefit-sanctions-health
 
Description The research team has focussed on developing relationships with policy and data analysts from public sector organisations including Scottish Government, NHS Health Scotland, the Department for Work and Pensions, Her Majesty's Revenue and Customs, the Office for National Statistics and advocate organisations, including Joseph Rowntree Foundation and Child Poverty Action Group. We have discussed both interim findings and plans of research activities at a range of workshops and meetings with the view of exploring ways in which our research may contribute to the research and policy activities of relevant non-academic organisations and identify avenues for research collaboration. This award supported capacity building activities undertaken by the early career researcher within the team. The employed early career researcher attended a summer school in order to enhance her analytical skills in the area of causal inference. She has also provided training as part of a course organised by the Scottish Centre for Administrative Data Research, addressed to social and health researchers from both academic and non-academic institutions who are working with or interested in using linked administrative data for their own research.
First Year Of Impact 2019
Sector Communities and Social Services/Policy,Other
Impact Types Societal,Policy & public services

 
Description 'Devolved benefits in Scotland: policy, evaluation and rollout', Scottish Government Edinburgh, organised by Policy Scotland, University of Glasgow and Carnegie UK Trust 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact The event included presentations from researchers from Scottish Government and academia, as well as advisors working in the area of evaluation of devolved benefits. The presentation were followed by reflection and discussion among attendees and raised increased interest in our research project and related research activities.
Year(s) Of Engagement Activity 2019
 
Description 'Welfare reform in Scotland: research, policy and the third sector', organised by Policy Scotland, University of Glasgow 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach Regional
Primary Audience Professional Practitioners
Results and Impact This seminar included presentations from Nick Bailey, academic colleagues and researchers from Scottish Government. The event aimed to interrogate ways in which academic and partners can impact the developing of the social security agenda in Scotland and how new research collaboration can improve the evidence base for the social security system. The event led to increased interest in our research project and related research activities.
Year(s) Of Engagement Activity 2019
URL https://policyscotland.gla.ac.uk/welfare-reform-in-scotland-research-policy-and-the-third-sector/
 
Description Blog - Can we use linked administrative data to identify social disadvantage? 
Form Of Engagement Activity Engagement focused website, blog or social media channel
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Public/other audiences
Results and Impact This blog was written to share reflections about the research conducted for the following publication: Pattaro, S., Bailey, N., and Dibben, C. (2020) Using linked longitudinal administratrive data to identify social disadvantage, Social Indicators Research, 147(3): 865-895 (https://link.springer.com/article/10.1007/s11205-019-02173-1). The blog also offered some reflections on working with limitations of linked administrative data.
Year(s) Of Engagement Activity 2020
URL https://www.scadr.ac.uk/news-and-events/blog-can-we-use-linked-administrative-data-identify-social-d...
 
Description DWP's Areas of Research Interest, University of Glasgow 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact The workshop aimed at discussing findings from the research project and other research conducted by colleagues at the University of Glasgow. It also aimed at exploring potential avenues of research collaboration and engagement with colleagues from DWP Labour Market Analysis Unit and related units.
Year(s) Of Engagement Activity 2019
 
Description HMRC RTI and Self-Assessment Data Research Hack Day, organised by Office for National Statistics (ONS), Newport 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact The event aimed to identify products, concepts and research questions to enable public benefit research and improved statistical products. The event raised questions and a discussion with colleagues from academia, ONS and HMRC.
Year(s) Of Engagement Activity 2019
 
Description Health and inequalities in an era of crisis organised by the Centre for Research on Environment, Society and Health, Global Health Policy Unit at the University of Edinburgh and the Scottish Graduate School of Social Science - Doctoral Training Partnership 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Other audiences
Results and Impact This event explored the implications of the recent economic recession and austerity, rising tension over immigration and right-wing populism for health and health inequalities. The event was opened to researchers, policy-makers and advocates. It sparked questions and increased interest from both academic researchers and Scottish Government analysts.
Year(s) Of Engagement Activity 2019
URL https://lgiu.org/event/one-day-symposium-health-and-inequalities-in-an-era-of-crises/
 
Description Introduction to using administrative data for social and health research, Edinburgh, co-organised by Scottish Centre for Administrative Data Research (SCADR) and Scottish Longitudinal Study Development and Support Unit (SLS-DSU) 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact The training event was addressed to social and health researchers from both academia and public sector organisations working with linked administrative data. The presentations concerning the project and related research activities sparked questions, discussion and request for further information.
Year(s) Of Engagement Activity 2019
URL https://www.ncrm.ac.uk/training/show.php?article=10284
 
Description Making data research ready, pre-conference workshop, organised by 4th International Conference on Administrative Data Research, Cardiff 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach International
Primary Audience Professional Practitioners
Results and Impact The workshop was co-organised by Administrative Data Research-UK (ADR-UK) and Office for National Statistics (ONS); the audience included academic researchers working with administrative data, policy and data analysts from public sector organisations involved as partners in ADR-UK. The workshop sparked questions and discussion which led to increased interest from colleagues from academia and the various public sector organisations.
Year(s) Of Engagement Activity 2019
URL https://ijpds.org/adr2019
 
Description Meeting, organised by Head of Family Resources Survey, Surveys Branch, Department for Work and Pensions, Caxton House, London 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact The meeting involved colleagues from our research team, research collaborators from MRC/CSO Social and Public Health Sciences Unit, and leading officials from DWP (e.g. Family Resources Survey, Central Analysis and Science Strategy Unit, Labour Market, Families and Disadvantage Directorate). It involved discussion of initial research findings and exploration of avenues for research collaboration.
Year(s) Of Engagement Activity 2019
 
Description Scotland's DWP Areas of Research Interest Workshop, co-organised by Policy Scotland, University of Glasgow, Research Support Office, University of Edinburgh and DWP Central Analysis and Science Strategy Unit 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact The workshop brought together DWP analyst, Scottish Government officials and analysts, and leading Scottish academics working in the broad areas of research interest. The workshop created the opportunity for discussion of potential avenues for research collaboration and advancements in the area of employment and progression.
Year(s) Of Engagement Activity 2019
URL https://policyscotland.gla.ac.uk/scotlands-dwp-areas-of-research-interest-workshop-outcomes/
 
Description TalingData: ADR Scotland mini-summit, Edinburgh 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Professional Practitioners
Results and Impact This event included brief presentations and discussion on activity from across the ADR-Scotland partnership; it was addressed to colleagues from Scottish Government, Scottish Centre for Administrative Data Research (SCADR), electronic Data Research and Innovation Service (eDRIS), National Records of Scotland (NRS) and Edinburgh Parallelel Computing Centre (EPCC). The networking event sparked questions and interests from both academic colleagues and colleagues from Scottish Government, eDRIS and NRS.
Year(s) Of Engagement Activity 2019
URL https://www.scadr.ac.uk/news-and-events/talkingdata-adr-scotland-mini-summit
 
Description What's the next game changer for solving poverty in Scotland, organised by Joseph Rowntree Foundation, Edinburgh 
Form Of Engagement Activity Participation in an activity, workshop or similar
Part Of Official Scheme? No
Geographic Reach National
Primary Audience Policymakers/politicians
Results and Impact This event aimed to identify a plan of action to address the drivers of child poverty across work, housing and social security in Scotland. It included talks by executives of JRF and other third sector organisations, academic researchers and policy makers. The event included parallel workshop and participants raised questions and interests concerning our research project and related research activities.
Year(s) Of Engagement Activity 2019
URL https://www.jrf.org.uk/event/what%E2%80%99s-next-game-changer-solving-poverty-scotland