Does the past dictate the future? Exploring the impact of employability programmes on adults experiencing multiple disadvantage

ABSTRACT

The research explored the impact of an employability programme, delivered by a Community Interest Company and other third sector partners in England. The programme was designed to increase the employability of people aged between 16–72 years-old who were unemployed or economically inactive. To measure the impact of the programme on participants, 1,098 people engaging in the project completed questionnaires designed to capture demographic data and measure general self-efficacy (GSE) upon joining the programme (Time 1); whilst 163 of the same participants completed the questionnaire upon completing the programme (Time 2). Interviews were conducted with 26 participants engaged with the programme. Results of the questionnaire data analysis revealed a statistically significant relationship between levels of disadvantage experienced by the participants and GSE at Time 1; statistically significant increases in GSE levels between Time 1 and Time 2 for participants who completed the programme; and a statistically significant relationship between GSE at Time 2 and employment/training outcomes. Triangulation of the quantitative and qualitative results revealed the positive impact of the programme on participant’s self-efficacy and employability. This paper is the first of its kind in the UK to explore the impact of employability programmes on adults experiencing multiple disadvantage.

KEYWORDS:

• General self-efficacy
• unemployment
• outcomes
• disadvantage

Introduction

Between October 2018 and December 2019, an estimated 1.36 million people were classed as unemployed or economically inactive, 14,000 fewer than the previous quarter and 100,000 fewer than the previous year (ONS 2019). Despite the fact the United Kingdom has recently experienced a reduction in unemployment, there are still a significant number of people without meaningful employment. This paper explores the impact of an employability programme on unemployed or economically inactive individuals, with a specific focus on whether experiences of multiple disadvantage have a cognitive impact on self-efficacy. The paper begins by outlining the benefits of measuring the impact of the programme on participants before continuing with a discussion of the complexity of issues facing groups of unemployed or economically inactive people. Specifically, it considers the impact of multiple disadvantage on individuals and the role of ‘self-efficacy’ as a predictor of employability. Next, the methodological approach to the research is proposed, with emphasis on utilising a longitudinal questionnaire, as well as semi-structured interviews. Finally, this paper explores the long-term impact of the programme on participants, with recommendations for measuring this impact. The paper makes an original contribution to knowledge by highlighting the role that self-efficacy, and specifically General Self-efficacy (GSE) can have on an individual’s employability. In doing so it extends the prior literature, which has focused on the general unemployed (Eden and Aviram 1993), those not in employment, education or training (NEET) (Hazenberg, Seddon, and Denny 2012), or graduate unemployment (Hazenberg, Seddon, and Denny 2014). Whilst prior research has suggested that psycho-social constructs like self-efficacy can impact upon employability (Meyers and Houssemand 2010), this has not yet been explored on a large-scale in relation to adults experiencing multiple disadvantage in the UK. By exploring the role that the employability project has on augmenting GSE, this paper presents evidence for how future employment integration programmes can seek to support such individuals back into employment, whilst also improving their confidence, motivation and self-esteem (Judge, Locke, and Durham 1997). More specifically, this paper is the first of its kind in the UK to explore whether experiences of multiple disadvantage have a cognitive impact on self-efficacy.

Disadvantage and the impact on life experience

Disadvantage is a complex term with no commonly accepted definition across the globe. Traditionally disadvantage has been linked to income poverty; however, developments in understanding the multidimensional nature of disadvantage has resulted in identification of a range of variables (Saunders, Naidoo, and Griffiths 2007; Gordon et al. 2000). These variables move beyond income poverty to consider areas such as social exclusion (Dehaan and Maxwell 1998; Burchardt, Grand, and Piachaud 2002), access to public services, education and health. This multidimensional nature of disadvantage creates problems for researchers wishing to identify indicators of disadvantage in society. Considering the elements of disadvantage outlined by Saunders, Naidoo, and Griffiths (2007), Gordon et al. (2000), Dehaan and Maxwell (1998) and Burchardt, Grand, and Piachaud (2002), the indicators for deprivation identified by the Indices of Deprivation (IoD) may present a solution to identifying appropriate indicators. The indices of deprivation are categorised into seven domains (income, employment, education, health, crime, barriers to housing and living environment) with 39 indicators organised across these domains (Ministry of Housing, Communities and Local Government 2019). Within the IoD, indicators include income support families, claimants of job seekers allowance, remaining in education post 16 years-old, and English language proficiency.

The Index of Multiple Deprivation (Ministry of Housing, Communities and Local Government 2019) showed that 88% of neighbourhoods in England, had remained in the most deprived category since 2015. Incidences of deprivation in England are not isolated, with 61% of local authorities containing at least one neighbourhood in the most deprived category. Analysis of Index of Multiple Deprivation (Ministry of Housing, Communities and Local Government 2019) data shows that the 10% most deprived neighbourhoods in England are facing multiple disadvantage and challenge (Ministry of Housing, Communities and Local Government 2019). Indeed, around 98.7% of the 10% most deprived neighbourhoods are ranked most deprived in two or more domains and around 65.5% are ranked most deprived in four or more domains. This illustrates the fact that people living in the most deprived areas face multiple disadvantage (disadvantage in at least two domains) (Ministry of Housing, Communities and Local Government 2019). In England, 22.4% of the population live in the 10% most deprived neighbourhoods for income and 25% of the population live in the 10% most deprived neighbourhoods for employment. Links between employment and income deprivation mean that neighbourhoods ranked the most deprived for employment are ranked the most deprived for income (Ministry of Housing, Communities and Local Government 2019). Deprivation impacts on personal wellbeing, with individuals experiencing deprivations having higher levels of hopelessness, lower self-esteem, lower self-efficacy and poorer mental health (Packard et al. 2012).

Impact of unemployment

The terminology around employment is constantly changing, with Labour Market information from the ONS (2019) utilising an array of terms such as employed, unemployed, economically active, and economically inactive. Employment refers to people aged 16 years-old plus who engage in one hour or more of work per week; whereas unemployment refers to people aged 16 years-old plus who are without a job but have been actively seeking work in the past four weeks¹. Economically active refers to people aged 16 years-old and over who are either in employment or actively seeking employment whereas economically inactive refers to people aged 16 years-old and over without a job who have not sought work in the last four weeks and/or are unable to commence work in the next two weeks. Due to the nature and aims of the programme, this research uses the terms unemployed and economically inactive to describe participants engaged on the programme. Information from the Labour Force Survey (ONS 2019) illustrate a fall in the number of people unemployed and the number of people aged from 16 to 64 years-old who are economically inactive. In the United Kingdom, an estimated 32.6 million people were employed between October and December 2018, 167,00 more than the previous quarter and 444,000 more than the previous year (ONS 2019). This means that an estimated 1.36 million people are unemployed, 14,000 fewer than the previous quarter and 100,000 fewer than the previous year (ONS 2019). These unemployment figures are based on individuals not currently employed and whom are actively seeking and available for work, which means an estimated 8.63 million people aged 16 to 64 years-old are economically inactive (ONS 2019). This information shows that 75.8% of all people aged 16 to 64 years-old were in employment (ONS 2019). Overall, the unemployment rate was estimated at 4% in the United Kingdom, the lowest reported estimates since February 1975 (ONS 2019).

Despite the fact the United Kingdom has experienced a reduction in unemployment, there are a significant number of people with a desire to work (1.36 million individuals), who are unable to gain appropriate and meaningful employment (ONS 2019). The psychological impact of unemployment is well documented, with researchers (Jahoda, Lazarsfeld, and Zeisel 1972; Creed, Muller, and Machin 2000; Karsten and Moser 2009) finding that unemployment impacts on depression, anxiety, hopelessness, self-esteem, confidence, and self-efficacy. Research conducted by Karsten and Moser (2009) found that psychological issues were higher amongst the unemployed (34%) than amongst the employed (16%). Individuals experiencing unemployment or economic inactivity also experience stigma and negative stereotypes (Danckert 2017), which adds to the challenge of securing employment. This can impact on health, wellbeing and life satisfaction, with people experiencing unemployment and economic activity often ‘othered’ by society.

Reducing the negative impact of unemployment then, is reliant on reducing the stigma and negative stereotypes associated with unemployment or economic activity whilst improving health, wellbeing and life satisfaction, to guide people towards positive outcomes. Unemployment is a complex area, with individuals experiencing unemployment often experiencing other issues such as depression, anxiety, subjective well-being and self-esteem (Murphy and Athanasou 1999; Karsten and Moser 2009). Individuals experiencing multiple disadvantage will have different coping mechanisms to others, which will play a role in the psychological impact of unemployment. Determining whether individuals experiencing multiple disadvantage also experience psychological impact differently from others is central to this paper.

General self-efficacy as a predictor of employability

Research shows that positive experiences, success in life, confidence executing actions and psychological wellbeing augment self-efficacy (Seddon, Hazenberg, and Denny 2013). Self-efficacy is built through mastery experiences, the process of one trying and succeeding at tasks; and vicarious experience, the process of observing others you know succeeding at said tasks (Bandura 1977). Self-efficacy is a psycho-social construct based within social cognitive theory (Bandura 1977), which has been linked to employability amongst a wide-variety of different social groups, including the young and disadvantaged i.e. those not in employment, education or training (NEET) (Hazenberg, Seddon, and Denny 2012). Whilst self-efficacy can be related to very specific tasks, general self-efficacy (GSE) is a generalised construct related to an individual’s perceived ability to perform in everyday life (Schwarzer and Jerusalem 1995). Indeed, research has identified that self-efficacy is a generalised trait that affects individuals’ ability to perform in general life situations (Eden 1988; Eden and Kinnar 1991). GSE has also been linked to confidence, motivation and self-esteem (Judge, Locke, and Durham 1997) and can therefore be seen as a good indicator of an individual’s determination to persevere with tasks despite potential failures or setbacks (Tipton and Worthington 1984).

When exploring the employability of disadvantaged individuals, the role that psychological constructs such as GSE can play in predicting job-seeking behaviour and positive outcomes (such as securing employment) can be significant (Meyers and Houssemand 2010). Participation in education and training opportunities such as work-integrated learning programmes helps to augment self-efficacy, enabling individuals to obtain the necessary skills required for work (Bowen and Drysdale 2017; Smigiel, Macleod, and Stephenson 2015). Hazenberg, Seddon, and Denny (2012) identified the role that GSE, along with self-regulative efficacy and social self-efficacy, had with disadvantaged NEETs engaged in employability enhancement programmes; whilst prior research has also shown the impact that employability interventions, including those delivered through community programmes, can have on an individual’s GSE and job-search intensity (Eden and Aviram 1993; Creed, Bloxsome, and Johnston 2001). In relation to the current research, this is particularly pertinent, as the focus is on an employment integration programme delivered by a partnership of several community-based support organisations in a region in England. This support is delivered to a wide-range of unemployed individuals, many of whom have experienced multiple disadvantage and long-term unemployment. This research therefore seeks to explore the impact that the programme has on employability, positive employment and training outcomes for disadvantaged individuals that engage with the programme. It also seeks to explore whether there is a relationship between multiple disadvantage and self-efficacy amongst these individuals. In doing so, the paper seeks to identify the augmentation of general self-efficacy as a key determinant of positive employment outcomes for severely disadvantaged individuals, and hence seeks to focus programme designers and policy-makers on the need to ensure that employability support is centred upon this theoretical construct.

Research methodology

Research AIM

This research aims to reveal and compare the impact of engagement with the programme on individuals classed as unemployed or economically inactive. Expanding this, the research sought to explore whether experiencing multiple disadvantage has an impact on self-efficacy. The employability programme, delivered by a Community Interest Company and other third sector partners in the East Midlands, was aimed at participants who were seeking employment, and was introduced in 2016. The programme was designed to increase the employability of people aged between 16 years-old and 75 years-old who were unemployed or economically inactive, with emphasis on supporting participants experiencing long-term unemployment and ex-offenders. It was funded by the European Social Fund and the National Lottery Community Fund, offering work-based experiential learning and qualifications which are designed to enable participants to learn the skills required for work in specific sectors including logistics, childcare and security. Specifically, participants learn Warehouse and Mechanical Handling, Forklift Driving, Ofsted approved Childminding, and other pre-employment work training (i.e. CSCS Green Card). Participants have access to further support in obtaining ID (i.e. passport and birth certificate), driving lessons, PPE, food vouchers, C.V. writing, and transport (i.e. bicycle and/or travel vouchers). The length of the programme varies for participants, depending on the training and qualification offered, with participants on the programme for approximately 3–6 months.

The initial phase utilised a quantitative methodology with questionnaires completed at the beginning (T1) of participants engagement with the programme and for a second-time (T2) at a point approximately 6 months post T1, depending upon when their engagement with ended.² The questionnaires utilised the GSE scale (Schwarzer and Jerusalem 1995), 10-items on a 4-point Likert scale (responses from 1-not true at all, 4-exactly true). This allowed for comparisons between GSE on commencing (T1) and completing (T2) the programme. This phase of research aimed to test the following hypotheses:

Hypothesis One: Participants who experience multiple disadvantage will have a lower GSE level at T1 than participants who do not experience multiple disadvantage.

Hypothesis Two: Participants taking part in the employability programme, will display a statistically significant greater increase in levels of GSE from T1 to T2.

Hypothesis Three: Participant levels of GSE at T2 will demonstrate a statistically significant relationship with positive employment/training outcomes.

Hypothesis Four: Experiences of multiple disadvantage prior to engaging in the programme will be positively inversely related to positive employment/training outcomes.

After completing the questionnaires, participants were invited to participate in a semi-structured interview. Semi-structured interviews were conducted to contextualise finding and will enable causality to be posited (McLeod 1994).

Many of the participants were unwilling to participate due to personal circumstances, and in the end 26 participants engaged in interviews. The semi-structured interviews were designed to gather information on participant’s perceptions of the programme, with a chance to reflect on their journey through the programme. These semi-structured interviews were recorded and transcribed to allow for rigorous analysis.

Participants

A total of 1,098 participants completed the self-reported questionnaire and 26 of these individuals also participating in semi-structured interviews. Among the questionnaire participants, 847 (77.1%) were male, 228 (20.8%) were female, 2 were transgender (0.2%) and 21 (1.9%) did not disclose their gender. At the time of the intervention, the participants were aged between 16 and 72 years, with the average participant being 35.9 (SD = 13.1) years old. While participants originated from five major ethnic groups, the majority (82.4%) identified as White British. Most participants were educated up to the GCSE/O-Level grades A-C and professional/vocational equivalent (NVQ Level 2) level (35.2%), followed by those with no qualifications or qualifications below NVQ Level 1 (27.7%), and those educated up to the GCSE/O-Level Grades D-E and professional/vocational equivalent (NVQ Level 1) level (14.7%). This evidences that nearly four-fifths of participants (77.6%) did not progress beyond compulsory education, leaving school at aged 16 years.

Research has shown that academic achievements has a profound impact on self-efficacy which, in turn impacts on rates of employment and starting salaries (Walters and Zarifa 2008). Despite the fact most participants are educated up to GCSE/O-Level grades A-C and professional/vocational equivalent (NVQ Level 2) level, research has shown that individuals need to be motivated to apply the skills learned (Schunk 1991). Further, when the current dataset is explored³ with regards to academic attainment and GSE at Time 1, the data reveals those individuals with GCSE grades D-E or lower as their highest educational achievement scored −0.78 (−1.95%) lower in GSE than their peers with GCSE grades A-C or higher as their highest educational achievement. This offers support to Walters and Zarifa’s (2008) research that identified this link between educational attainment and self-efficacy. However, when this analysis was repeated for individuals based upon their changes in GSE between the beginning and the end of the programme, no statistically significant differences were found based upon education.

The length of the participants’ unemployment varied between one and 463 months with the average time a participant spent unemployed being 33.3 (SD = 56.2) months; whilst 76 (6.9%) of the participants declared not having claimed unemployment benefits, the rest have been claiming unemployment benefits from 1 to 420 months, with the average being 22.1 (SD = 44.4) months. The interview participants were not required to provide identifying information; however, the demographic breakdown was largely similar to the questionnaire. With regards to disadvantage and the focus within this paper on multiple disadvantage, the participants were also able to identify with seven disadvantage categories (as many as applied), including: physical disability; mental health challenges; learning difficulties; ex-offender; homelessness; unemployed; and economically inactive (i.e. carer, student, not working but not claiming benefits). The data for the sample in relation to disadvantage is displayed below in Figure 1 and demonstrates that the main disadvantage characteristics in this sample were unemployment (N = 880; 80.1%); ex-offender (N = 216; 19.7%); being economically inactive (N = 135; 12.3%); and having mental health challenges (N = 121; 11%).

Figure 1. Sample disadvantage characteristics.

These characteristics were also collated numerically per individual on a scale ranging from 1–7, depending upon how many of the characteristics an individual selected. This analysis reveals that 742 (67.5%) of the participants identified with one disadvantage characteristic or less; whilst 356 (42.5%) identified with two or more characteristics (two participants identified with five). This data was also categorised into a nominal variable of ‘Multiple Disadvantage’ and ‘No/Singular Disadvantage’ for analysis purposes in relation to the hypotheses. Postcode data was also collected from the participants at Time 1, with 731 individuals supplying this data. This reveals that whilst there was a range of individuals living in areas that ranked from IMD decile 1 (the top 10% most deprived) through to IMD decile 10 (the top 10% least deprived), the average for the cohort was 3.5. Indeed, when breaking this data down it shows that 48% of the individuals that provided post-code data lived in the top 20% most deprived areas (IMD ranks 1 and 2) with a further 22.6% living in IMD deciles 3–4. As disadvantaged areas are classed as IMD deciles 1–4, this means that by postcode 60.6% of the participants could be classed as disadvantaged.

Finally, of the 1,098 participants that engaged in the research at Time 1, 163 completed exit questionnaires at Time 2, giving a longitudinal retention rate of nearly 15%. Whilst this was not as high as the research team had hoped for, given the levels of disadvantage experienced by many of the participants, a high drop-out rate was to be expected. One-way ANOVAs were used to explore potential demographic differences (age; length of time unemployed; length of time on benefits; disadvantage characteristics frequency; and IMD decile), as well as GSE, between those that completed the longitudinal phase of the research and those that did not. No statistically significant differences were found between the two samples. Cross-tabulation Chi-square analysis was also conducted in relation to gender as a factor in longitudinal completion. This did reveal a statistically significant difference (N = 1,077; p< .01; X² = 10.2) with 20.4% of males completing the programme versus 11.4% of females. However, given that gender is not a focus of this paper this is not considered a bias issue in the analysis.

Procedure & measures

The programme participants were invited to complete the questionnaire at the beginning of their engagement with the programme (T1). The questionnaires were completed by the participants through the organisation delivering the programme and were collected by the researchers on completion. The participants then completed the questionnaire for a second-time (T2) at a point between 3–6 months’ post T1, depending upon when their engagement with the programme ended. The questionnaire included the General Self-efficacy scale (Schwarzer and Jerusalem 1995), which includes 10-statements that participants rat themselves against in relation to a 4-point Likert scale (responses from 1-not true at all, 4-exactly true). The questionnaire was refined in collaboration with the delivery organisation and participants were able to complete the questionnaire with key workers, if required, to aid understanding.

After completing the questionnaires, participants were invited to participate in a semi-structured interview. Many of the participants were unwilling to participate due to personal circumstances, and in the end 26 participants participated. The semi-structured interviews were designed to gather information on participant’s perceptions of the programme, with a chance to reflect on their journey through the programme. These semi-structured interviews were recorded and transcribed to allow for rigorous analysis.

Quantitative analysis

Questionnaire data were analysed using Statistical Package for the Social Sciences (SPSS) version 22.0. Univariate analysis was performed to outline the descriptive statistics for the variables explored in the study, whilst bivariate analyses was performed to explore the associations between the independent variable (multiple disadvantage) and the dependent variable (self-efficacy). Descriptive statistics were used to explore participant’s demographic background, whilst one-way ANOVAs were used to ascertain differences in T1 GSE between groups (multiple disadvantage versus non-multiple disadvantage) and the relationship between GSE at T2 and positive employment/training outcomes. Paired-sample t-tests were utilised to establish if there were any longitudinal changes in participant’s levels of self-efficacy between T1 and T2. Finally, cross-tabulation Chi-squared analysis was conducted to explore the between-groups relationships (multiple disadvantage versus non-multiple disadvantage) at T1 and positive employment/training outcomes. A total of 163 participants completed a longitudinal dataset (Time 1 and Time 2).⁴ In statistical terms, a sample of 163 is still relatively small, and so caution should be exercised when interpreting this data. The scale ranges are reported for the reader’s convenience.

It should be noted that the dichotomising of the sample on the basis of disadvantage (multiple disadvantage versus non-multiple disadvantage) is not without its limitations. Indeed, the procedure adopted was based upon a limited sample within which randomised, matched samples were not possible. The analysis can therefore not be one that is considered as part of a randomised control group as would be seen, for instance, within a Randomised Control Trial (RCT) methodology. Rather, the analysis reported in this paper is one that is based within a comparison group design. The comparative approach adopted in the current research therefore presents certain threats to the validity of the research that must be acknowledged. First, in terms of internal validity with regards to potential differences between the two groups embedded through selection (Rosenthal and Jacobson 1968). Second, with regards to external validity, whereby the generalisability of the findings to the wider population are questionable (Cook and Campbell 1979). Research has identified that non-randomised control groups can sometimes offer misleading results when delivered in the same experimental setting as a randomised control group (Deeks et al. 2003). However, the same authors also acknowledge that sometimes RCT and randomised sampling is not possible due to ethical and/or infeasible (ibid). Within the programme delivered and reported on in this paper, this was the case as it would have been unethical to deny participants access to the support, whilst the nature of the sampling frame was purposeful as we wished to compare multiple versus singular/no disadvantage within the same intervention programme.

Whilst comparison groups are not considered as robust as control groups, they do still offer the researcher the opportunity to conduct statistical analysis that identifies trends in the data and that can demonstrate causal change. When this is embedded within a mixed-methods approach as undertaken in this study, the quantitative data analysis can then be used to support the qualitative analysis through a process of triangulation (McLeod 1994). This allows for some of the limitations of a non-randomised control group to be overcome through multiple data-points. Nevertheless, it does mean that some caution should be utilised when interpreting the statistical findings reported in this paper.

Qualitative analysis

Semi-structured interviews were analysed using the ‘Constant Comparative Method’ (CCM) (Lincoln and Guba 1985). This method allows for an iterative process using a ‘Grounded Theory’ approach (Glaser and Strauss 1967), utilising the five fundamental stages outlined by writers including Yin (1989) and Miles and Huberman (1994). These stages are: immersion, categorisation, phenomenological reduction, triangulation and interpretation, all of which are essential in conducting CCM. This allows for inductive reasoning, focusing on the emergence of categories from data rather than utilising predetermined categories (Maykut and Morehouse 1994). The stages used to analyse transcripts from interviews (as mentioned above) are:

1. Immersion – identify the ‘units of analysis’;

2. Categorisation – ‘categories’ emerge from the ‘units of analysis’;

3. Phenomenological reduction – ‘themes’ emerge from the ‘categories’;

4. Triangulation, support for researcher interpretations of ‘themes’ is sought in additional data;

5. Interpretation of findings is conducted in relation to prior research and/or theoretical models (McLeod 1994).

During ‘immersion’, transcripts were repeatedly interrogated, identifying different ‘units of analysis’ from the data, for example, ‘unemployed’, ‘offending’, ‘boredom’, ‘consequences’ (see Appendix A). These ‘units of analysis’ were further explored in the ‘categorisation’ stage in which concepts with similar meanings were grouped together according to ‘rules of inclusion’. This process resulted in 14 ‘categories’ emerging from 48 ‘units of analysis’. This process was replicated for the ‘phenomenological reduction’ stage, with ‘categories’ grouped according to similarities, creating a ‘rule of inclusion’ that resulted in the emergence of four ‘themes’. It is essential to note that when a ‘unit of analysis’ or ‘category’ did not comply with the ‘rule of inclusion’ for an existing ‘category’ or ‘theme’, new ‘categories’ or ‘themes’ emerged with a new ‘rule of inclusion’. The four emerging ‘themes’ were subsequently interpreted as: ‘needs’, ‘personal development’, ‘empowerment and motivation’ and ‘resilience and supporting others’. A diagrammatic illustration of this qualitative analysis process is provided for further clarification (see Figure 2). The numbers presented in the Categories boxes relate to the specific Unit of Analysis included within that category, whilst the numbers in the Theme boxes correspond to the relevant Categories contained within that theme.

Figure 2. Phases of CCM analysis.

Results

Quantitative results

Statistical analysis was conducted on the sample in relation to the four hypotheses outlined earlier, and the results of this analysis are presented here. One-way ANOVAs, paired-sample t-tests and cross-tabulation Chi-squared tests were utilised in this analysis.

Hypothesis One: Participants who experience multiple disadvantage will have a lower GSE level at T1 than participants who do not experience multiple disadvantage.

A one-way ANOVA was utilised to explore the relationship between disadvantage (factor) and GSE (dependent variable). Individuals that identified with zero or one disadvantage characteristic only were coded as ‘No/Singular Disadvantage’, whilst those that identified with two or more disadvantage characteristics were coded as ‘Multiple Disadvantage’. The results revealed a statistically significant difference in GSE scores (N = 1,051; p< .05; F = 5.8), with individuals experiencing multiple disadvantage having on average 1.2 lower GSE (equivalent to 3%) than those experiencing no/singular disadvantage. Whilst the difference in scores is not large, for a trait like GSE that is considered to be very stable over time, it represents a significant difference when considering employability. Hypothesis One confirmed. Table 1 below outlines the findings:

Hypothesis Two: Participants taking part in the employability programme, will display a statistically significant greater increase in levels of GSE from T1 to T2.

Table 1. GSE scores at T1 with disadvantage as the factor.

Scale	Disadvantage	N	Scores at T1	SD	F
GSE	No/Singular Disadvantage	710	31.2	5.4	5.8*
	Multiple Disadvantage	341	30.0	5.2

NB. * = p< .05, ** = p< .01, *** = p< .001.

Paired-sample t-tests were utilised to explore changes in GSE over time for the 163 participants that completed the longitudinal element of the research. The analysis revealed that there was a statistically significant increase in GSE over time (N = 163; p< .001; t = 3.3), with participants that completed the longitudinal element of the research showing an average increase in GSE of +1.4 (equivalent to +3.5%). This is a similar increase in percentage terms to those shown in prior research with NEETs and unemployed graduates engaging with employability programmes (Hazenberg, Seddon, and Denny 2012, 2014). Hypothesis Two confirmed. Table 2 below outlines this data.

Hypothesis Three: Participant levels of GSE at T2 will demonstrate a statistically significant relationship with positive employment/training outcomes.

Table 2. Changes in participant GSE scores between T1 & T2.

Scale	Time	N	GSE Score	SD	t
GSE	Time 1	163	30.9	5.1	3.3***
	Time 2		32.3	4.3

NB. * = p< .05, ** = p< .01, *** = p< .001.

Participant positive employment and training outcomes were also monitored within the data collection. Data was collated in relation to outcomes for 95 of the 163 participants that completed the longitudinal element of the research. A positive outcome included gaining employment, gaining a qualification (Forklift Truck Licence or NVQ Level 1 in Warehousing); and entering into a volunteering placement. A negative outcome was recorded if the participant had left the programme unemployed, without any new qualifications or to enter into a carer role with family.⁵ The data here revealed that of the 95 individuals for whom outcomes and GSE at T2 data were collected, 63 positive outcomes were identified (34 individuals gained employment; 26 individual gained a Forklift Truck Licence; 3 individuals gained an NVQ Level 1 in Warehousing), whilst 32 individuals experienced negative outcomes (i.e. remained unemployed and economically inactive).

A one-way ANOVA was employed to explore the relationship employment outcomes (factor) and GSE (dependent variable). The results revealed a statistically significant difference in GSE scores (N = 95; p< .05; F = 4.5), with individuals experiencing positive employment/training outcomes having on average +2.1 higher GSE (equivalent to 5.3%) than those experiencing negative outcomes. Again, given that a trait like GSE is considered to be very stable over time, it represents a significant difference and suggests that higher levels of GSE can contribute to improved employability. Hypothesis Three confirmed. Table 3 below outlines the findings:

Hypothesis Four: Experiences of multiple disadvantage prior to engaging in the programme will be inversely related to positive employment/training outcomes.

Table 3. GSE scores at T2 with employment/training outcomes as the factor.

Scale	Employment/Training Outcome	N	Scores at T2	SD	F
GSE	Negative Outcome	32	30.7	4.0	4.5*
	Positive Outcome	63	32.8	4.7

NB. * = p< .05, ** = p< .01, *** = p< .001.

A Cross-tabulation Pearson Chi-squared test was run to explore the relationship between the binary nominal variables of ‘No/Singular Disadvantage versus Multiple Disadvantage’ and ‘Positive/Negative Employment/Training Outcomes’. The analysis revealed no statistically significant relationship between the two variables. This suggests that prior levels of disadvantage were not a factor that impacted employability post-completion of the employment enhancement programme. Hypothesis Four not confirmed. In addition to this analysis, the research team also sought to explore whether educational attainment was related. A Cross-tabulation Pearson Chi-squared test was run to explore the relationship between the binary nominal variables of ‘Prior Educational Attainment’⁶ and ‘Positive/Negative Employment/Training Outcomes’. The analysis revealed no statistically significant relationship between the two variables. This suggests that prior levels of education were not a factor that impacted employability post-completion of the employment enhancement programme.

Qualitative results

Analysis was conducted for the interviews, with four themes emerging: ‘needs’, ‘personal development’, ‘empowerment and motivation’ and ‘resilience and supporting others’ (see Figure 2).

Theme A: needs

Experiencing disadvantage and deprivation can impact negatively on mental health (Packard et al. 2012), self-esteem and self-efficacy (Bandura 1977). This can result in uncertainty, which leads to challenges in accessing and/or maintaining suitable employment. Many participants experienced multiple disadvantage ranging from drug misuse and criminality, through to homelessness and domestic violence:

It was a horrible life … . I was drug-dealing, I was driving cars without licence, I was drinking, I was taking drugs all the time, I’ve never had any real friends they were all idiots, just a waste of time … I got into that at a very young age, so I didn’t know any different … childhood I had, parents that were drinking and taking drugs, having people around, watching junkies in the chair going to sleep, because there were needles hanging from their arms, things like that … I’d seen people die, I’d seen people overdose, I’d seen people getting hit by cars, I’d seen people being beaten up with a fucking baseball bats, that’s been horrible … . (P12)

I lost my children through domestic violence … they’ve - the five years that they’ve been gone from me … I’ve been unemployed for a year … I weren’t there really when my Mum was ill and then we lost her and then it was just like, I wasn’t there for her … And now the Dad’s not very well, he’s on end of life and it’s like, ‘Oh’, you know … You know, it’s just horrible. It’s just like, going through it all over again. (P20)

These experiences illustrate the multiple disadvantage participants face which impacts on employment, relationships, personal development and belief in oneself (Murphy and Athanasou 1999; Karsten and Moser 2009). These disadvantages are often complicated, with many participants struggling to find the resources necessary for improving their situation. Participants also discussed the role of the environment in promoting confidence and self-efficacy, with acknowledgement that education and training was key to them becoming employable:

Obviously it’s just the people that come here, it’s their lack of motivation. You know, some people that come here, they’ll do one day, and you never see them again … It’s the help really. The help that they offer and the training that they offer, it inspires you to do more. Like, you do one bit and you want to do more … Because just how it’s put together, how the people talk to you, how you are treated. It kind of makes you want to be here, it’s not the case of: I have to go to work today … I’m just at that point in my life where I thought: you know what, enough’s, enough … [without the programme] I think I’d be back in prison now. (P3)

Theme B: personal development

Experiencing multiple disadvantage has an impact on participant’s personal development. Personal development and change require belief in oneself that change can happen and trust in the others to provide the support required to achieve this change. Support for participants on the programme varied from one-to-one support through to group training. Interview participants discussed the process of personal development and change by highlighting the importance of support, confidence, belief, independence and improved well-being:

But when I came out of prison, looking at the situation I’d been in and where I need to go … So, for to find a place like this where you are not being judged on what you’ve done … feel a bit more respected and valuable … They’ve invested time and money to train me up. And then to take me on they’ve further invested in giving me work and contract … You come in here to give you a chance in life moving forward. (P6)

The most common change mentioned by participants on the programme was self-efficacy,⁷ enhanced psychological well-being, hopefulness, positive outlook for the future and the development of trusting relationships. Research shows that self-efficacy is a generalised trait that is linked to confidence, motivation and self-esteem (Judge, Locke, and Durham 1997) and can therefore be a good indicator of an individual’s determination to change. This determination is evident from this interview, when the trust and support of the lead delivery partner has led the individual to believe that they can overcome their multiple disadvantage:

… it sort of built this confidence in that somebody can give you a chance, so just confidence in looking for work, because [lead delivery partner] wanted to give somebody a chance. After completing a course and getting sort of certificates and qualifications for a forklift truck and warehousing, I stayed there a little bit longer volunteering a bit more. And the managers obviously saw that I’ve got something about me that they obviously wanted to keep. So, I was offered a full-time job … somebody is asking me to do it because they think that I’m capable of doing it. Again, it just builds that confidence that somebody has a bit of trust in you, they think you’re capable of doing it … (P14)

Although personal development for participants varies, the most common areas of change were self-efficacy, wellbeing, satisfaction with life, positive self-image, relationships, and motivation. Another area that helps empower participants to change, contributing to personal development, relates to participants feeling that regardless of past experiences (for example, involvement in criminal activity and/or substance misuse) there are employment opportunities in reach. This highlights participants’ determination to persevere with tasks despite potential failures or setbacks (Tipton and Worthington 1984).

Theme C: empowerment and motivation

Empowerment and motivation are referred to by participants in explaining their determination to gain employment. Many participants found that participating in the programme helped to trigger an understanding of their own employment-related needs, which empowered and motivated them to take control over their personal development:

I have done lots more networking talking to people and such and telling them what I’m interested in and it has helped me with my CV. It is helping me move forward and it has given me the right contacts so that is a relief … The fact that you can get information for your CV, I did not realise I had that many qualities that they have helped me realise that I have those qualities. (P1)

Developing strategies, knowledge and skills to achieve employment goals helps to motivate and empower participants to change. Motivations differed with some participants engaging out of necessity (for example, some individuals referred to the non-optional aspect of the programme), whilst others emphasised a desire and willingness to acquire new skills and qualifications. Indeed, the need and willingness to acquire new skills and qualifications was one of the underlying motivations for participants engaging in the programme. Another motivation, visible in the participant’s narrative, is the desire to change one’s life by obtaining a career or switching career path:

… so, I’ve been on the dole for about six or seven months. I was her carer, so I was not working, I was a cleaner, I felt like a maid, I kind of felt horrible to be honest like I should be doing so much more than being someone’s carer it was not the job for me … (P2)

Self-efficacy, described as one’s belief in their ability to accomplish tasks, has been identified as a critical factor in motivation and performance since Bandura’s (1977) original research. Participants on the programme have seen improvements in self-efficacy through improvements in cognitive skills, career choices, assertiveness, coping and performance which are all factors in securing employment (Bandura 1986; Murphy and Athanasou 1999; Karsten and Moser 2009).

Theme D: resilience and supporting others

Resilience relies on developing self-confidence and belief in one's capabilities, despite adverse experiences in the past (Simeon et al. 2007). Achieving the skills and qualifications to secure employment, then, helps individuals move from a world of negativity to a world of possibilities. This change creates resilience in individuals, which has a ripple effect on people around them and wider society:

I feel more [optimistic] – it gives me more things to go for [work opportunities] and it’s given me a lot more confidence in myself as all because I can do more than think I can … (P2)

Participants highlighted the benefits of the programme on self-esteem, confidence and employment.

… They’ve given me wages and a chance. So now I’ve got my own flat, I’ve got my car back. I’ve got a life back. My life now, I’ve got a girlfriend, I’ve enough money to pay my bills … But now I think about consequences of every action I do because every action there’s something at the end of it. (P6)

Life changes differed for individuals, with some participants experiencing dramatic changes in their lives (for example, obtaining employment, secure attachments) and others experiencing less dramatic, but equally positive changes (for example, attending the programme in of itself).

Discussion

Participants on the programme experience a multitude of disadvantage including substance misuse, homelessness, dysfunctional families, emotional/behaviour problems, health problems and involvement in criminality. Research shows that experiencing disadvantage can impact on an individual’s ability to secure employment (Murphy and Athanasou 1999; Karsten and Moser 2009; Danckert 2017). Results from the quantitative analysis illustrate that participants who experience multiple disadvantage have a lower GSE level on starting the programme (T1) than participants who do not experience multiple disadvantage. Analysis of the qualitative data also revealed augmentation of confidence, motivation and self-esteem. GSE has been linked to confidence, motivation and self-esteem (Judge, Locke, and Durham 1997) and can therefore be a good indicator of an individual’s determination to persevere with tasks despite potential failures or setbacks (Tipton and Worthington 1984). Individuals experiencing multiple disadvantage have additional obstacles that create the potential for failure and/or setbacks. Although experiencing one area of disadvantage can impact negatively on GSE levels, the impact is higher for those experiencing multiple disadvantage.

Understanding the impact of programmes on individuals experiencing disadvantage is essential in developing effective and sustainable services. Participants engaging in the employability programme displayed statistically significant increases in levels of GSE from T1 to T2, illustrating the programme’s ability to help participants to perform in everyday life (Schwarzer and Jerusalem 1995). Triangulation of qualitative and quantitative results show the beneficial impact of the programme on participant’s self-efficacy levels, with significant increases evident in self-efficacy accompanied by participants perceived improvement in wellbeing, confidence and self-esteem. These improvements move beyond impact on the individual to include an impact on family life, suggesting that the programme also has a beneficial impact on participant’s families. Overall, findings illustrate the importance of self-efficacy on motivation, personal development, family life and wellbeing, which are all important factors in securing employment (Bandura 1986; Murphy and Athanasou 1999; Karsten and Moser 2009). The impact of the programme on GSE is evident; however, understanding the impact of this on employment is essential for understanding the impact of the programme. Furthermore, given the multiple disadvantages many participants face, it is essential that the environment within the programme fosters the trust and support that aids individual change (Murphy and Athanasou 1999; Karsten and Moser 2009). This often means changing the long-held negative attitudes that are promoted through societal perceptions of unemployed people (Danckert 2017), whilst providing support.

Research (Jahoda, Lazarsfeld, and Zeisel 1972; Creed, Muller, and Machin 2000; Karsten and Moser 2009) shows that unemployment impacts on depression, anxiety, hopelessness, self-esteem, confidence, and self-efficacy. Research shows that GSE can have a positive impact on employment (Eden and Aviram 1993; Creed, Bloxsome, and Johnston 2001; Hazenberg, Seddon, and Denny 2014). There was a statistically significant difference in GSE scores based on employability outcomes, with individuals experiencing positive employment/training outcomes reporting higher GSE levels than those individuals who did not. This demonstrates that increased GSE levels are related to positive employment/training outcomes. Experiencing higher GSE levels means that individuals have the confidence and experience to succeed in securing employment.

The data explored in this study suggests then that GSE offers a robust theoretical lens for understanding individuals’ experiences of multiple disadvantage, the impact of employability programmes on said individuals, and their likely employability post participation on such programmes. The research therefore echoes the results of prior research with different populations, notably Eden and Aviram (1993) and Hazenberg, Seddon, and Denny (2012) and extends these findings to demonstrate the impact of GSE amongst individuals experiencing multiple disadvantage. However, the paper does not seek to suggest that GSE is the sole determinant of employability, nor the sole means to reduce disadvantage. Indeed, given the chaotic lives experienced by many participants and the significant disadvantage prevalent in their lives, such an assertion would be simplistic. This is also demonstrated within the data, in that longitudinal data was only gained from 163 of the original 1098 participants engaged (a drop-out rate of over 85% across the programme). Rather, we suggest that programmes that seek to understand the complexity of their beneficiaries’ experiences, work to build secure attachments with participants, whilst providing support, education and training that augments GSE, have a higher probability of delivering positive employment outcomes. Research has shown that possessing knowledge and skills does not, on its own, improve outcomes and that motivation and confidence in applying knowledge and skills is key (Pintrich and De Groot 1990; Schunk 1991).

Figure 3 below illustrates the theoretical model developed within this paper, in relation to the design and implementation of employability programmes for individuals experiencing multiple disadvantage. It demonstrates that a specific pathway of support is required, that first focuses on building secure attachments with participants, before moving on to augment GSE through the building of confidence, motivation and self-esteem (Judge, Locke, and Durham 1997) via focusing on participants emotional/physical needs, personal development, empowerment and resilience. Prior literature has explored the role of attachment to place in the employment and training decisions of people in disadvantaged areas, demonstrating that attachment and social networks are critical in creating employment opportunities (White and Green 2011). This paper suggests that such attachments and networks can be built through multi-stakeholder partnerships delivering employability programmes, providing that trust is built between participants and their host organisation, whilst the wider partnership offers the social networks required to seek employment. Once attachment and trust have been built, the programme can then seek to augment GSE by developing a participant’s confidence, motivation and self-esteem (Judge, Locke, and Durham 1997) through a process of meeting physical needs, supporting personal development, empowering the individual and increasing their resilience. This process if successful leads to job-search and employment, following which the individual is likely to refer others to the programme through their networks and act to support/mentor others. If the individual does not gain employment, then it is critical that the programme continues to support their needs until employment is achieved. Indeed, when dealing with multiple disadvantage and long-term employment, the design of such programmes needs to recognise that employability and employment are not easily achieved outcomes.

Figure 3. Increasing employability for individuals experiencing multiple disadvantage.

Summary

This research shows that keeping in touch with and securing motivation/engagement from participants with complex needs, over prolonged periods, is challenging. Participants engaged on the programme have experienced a range of complex needs ranging from substance misuse and criminal behaviour to homelessness and mental ill-health. This resulted in a high drop-out rate (85.2%), with participants completing the initial questionnaire unable to complete the follow-up questionnaires. Nevertheless, the strength of the results reported in this paper suggests an association between employability and GSE amongst unemployed individuals experiencing multiple disadvantage. The data also suggests that in order to achieve positive employment outcomes with this demographic group, employability programmes need to seek to build emotional and place-based attachment with individuals, before augmenting their GSE through developing individual’s physical needs, personal development, empowerment and resilience. Whilst initial GSE levels at time 1 were mediated by prior educational attainment, this did not impact either changes in GSE over time nor the attainment of positive outcomes for the participants. This offers interesting avenues for policy-makers and practitioners in understanding how to design and deliver more effective employability programmes.

Future evaluations of employability programmes may benefit from implementing questionnaires with self-efficacy scales in combination with semi-structured interviews to identify the outcomes and outputs. Despite the appropriateness of the methodology employed, it is recommended that caution is used in interpreting the results as the questionnaire analysis was based on the comparison of a large Time 1 sample with a smaller combined longitudinal sample. Indeed, this means that there is a bias in our sample in that we only explore the perceptions of GSE changes of a small proportion of participants (i.e. those that complete the programme). This is further complicated by the fact that the comparisons made in this research between those individuals classed as experiencing multiple disadvantage and those classed as not (or only experiencing singular disadvantage), were not part of a full RCT with randomly matched samples Therefore, further research to explore outcomes for those individuals with multiple disadvantage is therefore required to better understand needs and impact.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Goodwill Solutions CIC.

Notes on contributors

Claire Paterson-Young

Claire - Dr Claire Paterson-Young (BA MSc PhD) is an Associate Professor at the Institute for Social Innovation and Impact (ISII). Her research focuses on the development of social impact measurement frameworks that aid organisations to identify the intended and unintended impact of interventions. Claire is on the editorial board for the Journal of Qualitative Criminology and Criminal Justice and Associate Editor for the Journal of Child and Family Studies.

Richard Hazenberg

Richard - Professor Richard Hazenberg (BA MA PhD) is the Director and Research Leader of the Institute for Social Innovation and Impact at the University of Northampton. He has contributed to international/national government policy through papers, conferences and roundtable meetings (including for the European Commission; OECD; Cabinet Office; and HM Treasury). Professor Hazenberg is on the editorial board of the Social Enterprise Journal and is a reviewer for a number of international peer-review journals. Professor Hazenberg is the University’s leading researcher for social innovation and impact.

Notes

1. Individuals also need to be available to commence work in the next two weeks, and/or people out of work who have found a job and are waiting to start in the next two weeks.

2. This engagement could end due to participant drop-out or an individual entering employment or formal training.

3. The data was dichotomised into those individuals with GCSE grades D-E or lower as their highest educational achievement, and those individuals with GCSE grades A-C or higher as their highest educational achievement. One-way ANOVAs were then run with respect to this binary educational variable and GSE scores at Time 1 and changes in GSE scores between the beginning and the end of the intervention.

4. Some statistical tests do not include the full 163 participants here, as some of the longitudinal questionnaires had missing question responses and so could not be used in the overall analysis.

5. It should be noted that the research team make no pejorative judgment of these outcomes for individuals. Indeed, suspending seeking employment to care for a relative can be viewed in the familial setting as a positive outcome. Furthermore, given the levels of disadvantage experienced by many participants, completing the programme itself and continuing with job-search can also be classed as a positive outcome. However, the purpose of this paper is to explore the impact of disadvantage on GSE and employability, hence a decision was taken to code only employment, training and volunteering outcomes as positive.

6. As noted earlier, the data was dichotomised into those individuals with GCSE grades D-E or lower as their highest educational achievement, and those individuals with GCSE grades A-C or higher as their highest educational achievement.

7. It should be noted that the participants themselves did not use the term ‘self-efficacy’, rather this was inferred by the research term through the process of phenomenological reduction in relation to participants discussing confidence, motivation and self-esteem.

Appendix A

Table

Units of Analysis

Mistake Consequences Mental Health Member of society Unemployed Chances Prison Offending Financial Issues Never-ending cycle Goals Family support Belonging Making a change Uncertainty Low self-esteem Lack of confidence Optimism Strength Nothing to do Boredom Better life Longing No trust

(25) Choices (26) Escape (27) Negative self-image (28) Desire to succeed (29) Regrets (30) Hopelessness (31) Medication (32) Control (33) Homelessness (34) Anxiety (35) Substance Use (36) Mess (37) Motivation (38) Additional support needs (39) Wasted time (40) Violence (41) Trauma (42) Domestic violence (43) Family sickness (44) Barriers (45) Criminal record (46) Benefits (47) Temporary job (48) Develop myself