Skip to main content

Looking beyond access: How do graduate perceptions of their work vary by socioeconomic disadvantage?

Further work on our measure of disadvantage revealed an error in the generation of HESA measure deciles. Our output area files for England, Wales and Scotland contained statistics for higher level geographies (either local authorities, regions and/or countries), which had not been removed prior to the formation of the deciles.

HESA measure deciles have been recreated based on a total of 232,296 output areas (181,408 in England and Wales, 46,531 in Scotland and 4,537 in Northern Ireland). Around 1% of output areas changed from quintile 1 to a higher quintile or vice versa. Approximately 5% of output areas were affected when undertaking an analysis by decile. We have found the impact of this to be minimal and the conclusions of our research are not materially altered.

2022-05-23

Research into equal opportunity in higher education has tended to focus on access. In this insight, HESA researchers explore the relationship between socioeconomic disadvantage and graduate outcomes.

1. Introduction

To date (including in our publication earlier this month[1]), analysis relating to our measure of disadvantage[2] has concentrated on matters relating to access to higher education. In this piece, we switch our attention for the first time to outcomes. Specifically, we explore whether graduate views on the design and nature of their employment differ based on the socioeconomic background of the individual. Furthermore, we go on to suggest possible explanations for the findings we observe by looking at the interplay between these two factors and Standard Occupational Classification (SOC) 2020.

We hope that this study can advance knowledge in this field and support policymakers in each of the four nations, as well as staff within providers who assist those from less privileged backgrounds throughout their journey on a degree course.

2. Why does this matter?

 
2.1. National policy

Promoting equal opportunity for all is a key policy aspiration across the whole of the UK. For example, in England, this was highlighted in the Queen’s Speech in 2021, which noted that the ‘levelling-up’ agenda will involve widening opportunity within society.[3] In Scotland, the 2016 Fair Work Framework states that fair work should be available to all.[4] The Well-being of Future Generations (Wales) Act 2015 outlines one of the goals to be the establishment of a more equal country, where all have access to decent work.[5] In Northern Ireland, an aspiration of the restored executive, as detailed in a ‘New decade, new approach’, is the creation of an economy with opportunities for all.[6]

2.2. Higher education policy

In line with national priorities, a shared objective across the UK higher education sector is to ensure that there is equal opportunity for all, with each nation noting the need to support those from deprived backgrounds. In November 2021, Michelle Donelan (the Higher and Further Education Minister) announced that providers in England would be asked to revise their Access and Participation Plans, with greater focus expected to be placed on the destinations of disadvantaged students following graduation. In her speech at a Times Higher Education Campus Live event, she stated that ‘a student’s outcome after university needs to be as important to providers as a student’s grades before university’.[7] In Scotland, Recommendation 33 of the ‘Blueprint for Fairness’ report by the Commission on Widening Access states that the Commissioner for Fair Access should ‘consider what further work is required to support equal outcomes after study for those from disadvantaged backgrounds’.[8] The Tertiary Education and Research (Wales) Bill anticipates one of the roles for the newly formed Commission in Wales will be to support disadvantaged students into employment after they complete their studies.[9] The Department for Economy in Northern Ireland has highlighted that ‘higher education and the opportunities that it brings should be available to all, regardless of their background’.[10]

3. Data

Our sample consists of UK domiciled full-time first degree qualifiers from the 2017/18 and 2018/19 academic years. We additionally restrict the dataset to graduates who reported that their main activity after graduation was paid work for an employer in the UK and who responded to all three graduate voice questions. The final sample comprised of 169,690 individuals and is the same as that utilised in our previous insight on the association between SOC 2020 and the design/nature of work measure.[11] We split our UK-wide measure of disadvantage into deciles[2], with decile 1 representing the 10% of output areas in the UK with the highest average proportion of residents with below level 4 qualifications/in occupations that fell within NSSEC groups 3 to 8. Consequently, decile 1 represents the most disadvantaged group, with decile 10 consisting of those individuals from areas experiencing the lowest levels of deprivation.

4. Results

 
4.1. What are the main trends?

The first thing we note from the figures in the main part of the briefing, as well as the additional tables supplied in the appendices, is that the relationship between socioeconomic disadvantage and the design/nature of work measure varies across the four nations.

In England, we generally observe that those from more deprived backgrounds report lower design/nature of work scores. As mentioned in our previous insight, overlapping confidence intervals will not necessarily mean that the differences between two groups are not statistically significant. When examining whether the scores between decile 1 and another decile were significant through formal hypothesis testing, we found this to be the case at the 5 percent level in all instances, aside from when we compared decile 1 with decile 2 or 3. Further exploration of the individual components that constitute the composite measure illustrates that this differential in England is predominantly driven by the disparity in responses by socioeconomic disadvantage for the extent to which ‘work fits with future plans’. Indeed, no clear pattern emerges for the ‘skills used in study’ or ‘work is meaningful’ variable.

Figure 1: Mean design and nature of work score by decile of measure of disadvantage (England)

Scatter plot for England - fair work score broadly correlates with decile of disadvantage.

Figure 2: Mean graduate voice scores by decile of measure of disadvantage (England)

Scatter plot for England - "work fit with future plans" correlates with disadvantage but other graduate voice scores vary much less between disadvantage deciles.

In both Scotland and Wales, the design and nature of work scores of those in decile 1 seem to lag behind those reported from other deciles. For both nations, if one were limiting their analysis to those in deciles 2 to 10 only, no obvious trend can be seen from the data. For Scotland, the design and nature of work score for those in decile 1 is significantly different when compared with any one of the other nine deciles. In Wales, the differences are not significant at the 5 percent level when decile 1 is compared with either decile 2, 4 or 9. Considering the individual variables that make up the design and nature of work measure, those in decile 1 report the lowest scores across all three variables in Scotland and Wales. The greatest discrepancies are seen with respect to the ‘work fits in with future plans’ variable.

Figure 3: Mean design and nature of work score by decile of measure of disadvantage (Wales)

Scatter plot for Wales - "work uses skills" and "work fits with plans" scores are noticeably lower for graduates from the most disadvantaged areas (decile 1)

Figure 4: Mean graduate voice scores by decile of measure of disadvantage (Wales)

Scatter plot for Wales - "work uses skills" and "work fits with plans" scores are noticeably lower for graduates from the most disadvantaged areas (decile 1)

Figure 5: Mean design and nature of work score by decile of measure of disadvantage (Scotland)

Scatter plot for Scotland - fair work score is noticeably lower for graduates from the most disadvantaged areas (decile 1)

Figure 6: Mean graduate voice scores by decile of measure of disadvantage (Scotland)

Scatter plot for Scotland - "work uses skills" and "work fits with plans" scores are noticeably lower for graduates from the most disadvantaged areas (decile 1)

For Northern Ireland, no significant differences are found at the 5 percent level when assessing the relationship between the design/nature of work measure and socioeconomic disadvantage, though the sample sizes available to us are smaller in this part of the UK.[12]

Figure 7: Mean design and nature of work score by decile of measure of disadvantage (Northern Ireland)

Scatter plot for Northern Ireland - No clear relationship between fair work score and and disadvantaged

Figure 8: Mean graduate voice scores by decile of measure of disadvantage (Northern Ireland)

Scatter plot for Northern Ireland - No clear relationship between graduate voice scores and and disadvantaged

4.2. What may explain these findings?

In our preceding insight, we illustrated the link between SOC 2020 and the design/nature of work measure. When we analyse the association between SOC 2020 and our measure of disadvantage in England, we find that those in decile 1 are less likely to be in occupations that score highly on the design/nature of work measure (e.g. professional occupations), while also having greater probability of being based in positions where scores are relatively lower (e.g. sales and customer service occupations). Indeed, these differences in occupation by socioeconomic disadvantage appear to be part of the reason behind the overall lower design/nature of work scores of those from disadvantaged backgrounds.

However, the identified relationship between SOC 2020 and socioeconomic disadvantage does leave one wondering why there is no evident linear connection between the ‘using skills learned during study in work’ variable and socioeconomic disadvantage within England.

The key to understanding this is to look specifically at the ‘using skills learned during study in work’ scores by socioeconomic disadvantage within (rather than across) particular occupation groups. For example, focusing on those graduates in professional occupations, we observe that individuals from the most disadvantaged decile (1) report an average score of 4.26 for this component of our composite measure. In comparison, graduates in the least deprived group (10) have a mean value of 4.00, with a clear downward trend emerging as we move through the deciles. That is, those from more disadvantaged backgrounds in professional occupations have higher values for the extent to which they use their study skills in their job relative to graduates in professional occupations from areas of lower deprivation. A similar pattern is also seen if we analyse those in associate professional occupations. As a result, the lack of any clear association is partly due to the impact of the lower proportion of graduates from disadvantaged backgrounds in professional/associate professional occupations being counteracted by the higher scores reported by those from lower decile groups who are based in these types of positions. No such relationship exists when focusing on the ‘work fits in with future plans’ variable.

In the case of the ‘work is meaningful’ variable, the lack of any obvious relationship with socioeconomic disadvantage is partially explained by the fact that we find a higher proportion of disadvantaged students are based within the caring, leisure and other service category, where the scores for ‘work is meaningful’ are higher than almost all other occupations.

In Scotland, the association between SOC 2020 and socioeconomic disadvantage is very similar to what we observe between the design/nature of work score and deprivation. That is, the outcomes of those individuals from decile 1 differ from graduates in other deciles. In this instance, we see that a lower proportion of individuals from decile 1 are based in either professional or associate professional occupations. Conversely, a higher percentage of graduates within decile 1 are employed in elementary/sales and customer service occupations. In Wales, though the association between SOC 2020 and socioeconomic disadvantage does not align as closely with the relationship between the design/nature of work score and deprivation, the general trend is also one of disadvantaged graduates being less likely to be employed in occupations that score highly in the design/nature of work variable. Consequently, SOC 2020 is again likely to play a role in explaining the differences we observe in Wales for the design/nature of work scores by socioeconomic disadvantage.

5. Next Steps

Over coming months, we shall continue to draw upon the responses we received to our consultation on the design/nature of work measure to explore how various factors relate to the design/nature of work measure. Future insights relating to this measure will therefore be published in due course. In the meantime, comments on this (or our previous) releases are most welcome and can be sent to [email protected].


[11] We exclude those domiciled/employed in Guernsey, Jersey or the Isle of Man or those for whom we cannot assign a specific region of domicile/employment in the UK. As the purpose of this insight is to advance knowledge on this issue within the sector, we limit the dataset to those providers that agreed for the data they had submitted to be used for such purposes. More detail on this can be found at https://www.hesa.ac.uk/support/provider-info/subscription/onward-use.

[12] Formal hypothesis testing involves conducting the t-test for differences in means between two independent samples. Please note that the findings around statistical significance discussed in this section apply whether one assumes equal or unequal population variances. Confidence intervals reported in the appendices are based on the t-distribution.

Share
Insight
Tej Nathwani

Tej Nathwani

Principal Researcher (Economist)
Siobhan Donnelly

Siobhan Donnelly

Lead Statistical Analyst

Contents

 

See more research from HESA

 

Sign up for Research releases