The Impact of Welfare Reform in Scotland - Tracking Study

3. Methodology

• The study utilises a qualitative longitudinal approach in order to best track participants' experiences over time, as the welfare changes are introduced.
• All participants interviewed are currently in receipt of working age welfare benefits, and were selected using a purposive sampling strategy which was designed to reach those in receipt of benefits from across Scotland, and cover a diverse set of household circumstances.
• In-depth, semi-structured interviews were carried out with 43 participants.
• The data was transcribed and analysed using the qualitative analysis computer software NVivo.

3.1 The study takes a qualitative longitudinal approach, which involves interviewing participants six times over three years to 2016, with the first two sweeps of interviews being started in September 2013 and April/May 2014. The longitudinal approach allows changes to be mapped out over time, as they occur. The study design also allows issues to be explored at the time that participants present them.

3.2 Longitudinal social research, which typically involves in-depth interviews with participants who are re-interviewed across time in order to map any changes in their circumstances, is particularly useful for a study exploring the impacts of welfare reform. By returning to participants, we are able to give them the opportunity to reflect on any changes that might have occurred since their last interview. It also allows the researcher to tailor individual interviews in order to find out more about specific sets of circumstances (Farrall, 1996). This is especially useful when interviewing people who are claiming benefits for a number of different reasons such as disability and/or because they are a jobseeker and/or lone parent. The longitudinal nature of the research can also capture the experiences of participants at different stages of welfare reform as it is rolled out, and to identify other changes, including those in their labour market status and how these have affected them.

3.3 The first sweep of interviews, which is presented in this report, is designed to gather initial baseline data about the participants and their households as well as the core longitudinal questions.

Typical profile of participants

3.4 Participants had a variety of different reasons for claiming benefits. Although the exact combination and amount received varied according to each household's individual circumstances, there are nonetheless a number of typical benefit combinations, and most participants fell into one of these categories.

• Lone parents whose youngest child has turned five years of age who are not in work and are typically in receipt of JSA as their wage replacement benefit, and are claiming HB and CTB. They also receive CB and CTC for their child(ren). Child benefit is received for each child, although a lower rate is paid for any additional children after the first born. Prior to the youngest child turning 5 years of age, lone parents can claim IS instead of JSA. Two parents in the study were in receipt of IS, but all other parents, including lone parents, in the study had been moved onto JSA at time of interview.
• Full-time carers in receipt of IS and CA as their income replacement benefit, and by association HB and CTB, along with DLA for the person they are caring for.
• Single people and couples who are looking for work and are claiming JSA, either contribution or income based depending on their working history and the income of their partner, if present in the household. Whilst claiming JSA, they may also be entitled to claim HB and CTB.

Selecting and recruiting the sample

3.5 A purposive sampling strategy was used. The main criterion for inclusion in the sample was that the household had direct experience of at least one of the following benefits changes:

• Benefit cap
• Additional hours required for working tax credit
• Changes to lone parents' obligations
• Receipt of Universal Credit
• Disability Living Allowance
• Employment Support Allowance.

3.6 Within this, there is considerable scope for variation; for example, households affected by these changes may or may not have children, could be couple or single adult households, and be in or out of work. A sample framework of minimum estimated participant numbers for a range of different household characteristics guided recruitment. Data from the Labour Force Survey (Q1 2013) about the percentage of people in Scotland on JSA, ESA, IS and DLA provided guidance in the drafting of the sample framework. The sampling also sought to 'represent' spatial (e.g. rural -urban) factors that could potentially affect the outcomes of interest (see Table 3.1 for details of the sample characteristics, both the target sample size and the actual sample size). However, it should be acknowledged that this sample is not meant to be representative. This means that there are limitations in the conclusions that this study can provide in terms of the experiences of some groups of participants (e.g. lone parents, those living in rural areas). However, the study provides valuable insights into the experiences of those in receipt of benefits and highlights some of the issues faced by specific groups which could be followed up in more depth in other research.

3.7 Recruitment of participants primarily took place through gatekeeper organisations (i.e. approaching relevant third sector and public agencies who deal with the relevant groups); additional recruitment also took place through snowball sampling (where participants recommended potential participants for a study). This strategy was used as it not only made it easier to recruit suitable participants, but also to keep in touch with them for the follow up interviews.

3.8 After this initial recruitment the sample will be re-contacted prior to each repeat wave of fieldwork. Maintaining the interest of the sample should be thought of as an on-going process, rather than a one-off event. Therefore there will be some on-going contact, either directly with participants or through the gatekeepers, which should help to minimise sample attrition.

3.9 The required sample size at the end of the study is 30 households. However, the initial sample needed to be larger than this to allow for some attrition, as participants may be unwilling or unable to participate in future sweeps, or the research team may lose contact with them. It is difficult to estimate the degree to which attrition will occur. However, in order to reduce the risk of participants refusing to take part in subsequent interviews, where possible, the same researcher will return to conduct interviews in order to provide consistency for participants. The sample size for this first sweep is 43 (see Table 3.1).

Table 3.1: Overview of sample characteristics

Data collection

3.10 In-depth, semi-structured interviews lasting between 30 to 90 minutes were carried out with participants (see Appendix 2 for the schedule). In-depth interviews are seen as the most effective method as they are particularly suited to gathering data on individuals' personal histories, perspectives, and experiences. The interviews were semi-structured, with a brief and topic guide that gives a clear idea of the issues that should be covered, but with most questions focusing on open responses and the opportunity for the participant to raise other issues. This method allows participants to give rich, personal and in-depth accounts of their experiences, and also allowed for the researchers to build a rapport with participants (this was especially important in helping to minimise sample attrition between Sweeps 1 and 2). Interviews were conducted in a private setting in which participants felt comfortable, such as in their own home, or in a more neutral setting such as an advocacy organisation.

3.11 Participants were given an information sheet before participating in the study (see Appendix 4), and full consent was obtained before proceeding with the interview (see Appendix 3 for the consent form). Interviews were audio recorded where permission was given, and partially transcribed (i.e. relevant content from interviews, such as the households' accounts of their experiences, but not incidental conversation or 'warm up' questions).

3.12 The content of the baseline interview covered a number of areas (see Appendix 2 for the schedule):

• Financial information - level and sources of current income (including some cross-checking against expected entitlements), household costs and expenditure, budgeting and debt;
• Current employment and training/education;
• Sources of support - the use of services and formal support (and any barriers to accessing these), and informal support networks including family and friends;
• Household members' self-reported physical, mental and emotional health, and any use of healthcare services;
• Challenges people are currently facing, how they are coping and what is helping or hindering them;
• Other areas for discussion on welfare changes identified by participants, including participants' perspectives on welfare changes as a whole; and
• Retrospective evidence on income sources and employment over the previous two years.

3.13 No payment for time provided by participants was given. However, participants were given a voucher to compensate for travel, and any other expenses (including childcare), at a rate of £10 per household per meeting.

Analysis

3.14 The content of the interviews was analysed (including using qualitative analysis computer software NVivo)^[17] for important and/or recurring themes.

3.15 Although the sample in this research is small and qualitative, and any generalisation is difficult, it is still useful to observe whether patterns emerge in participants' experiences, not least to suggest avenues for future, larger scale research. However, this study also seeks to preserve narratives rather than reduce them to constituent parts, in order to understand people's individual experiences and the impact that policies will have on them.

3.16 This study also allows for the identification of key features of a participant's life as measured by established instruments (such as those used in national surveys to classify their type of accommodation or to measure of wellbeing), and to link these elements together, and situate the participant in their own unique context.

Research Ethics

3.17 This study received research ethics approval from Edinburgh Napier Business School's Research Integrity committee.