Skip to main content

Using Census data to generate a UK-wide measure of disadvantage - Next steps

6. Further remarks and next steps

This paper has supplied the rationale and methodology behind a new area-level measure of socioeconomic disadvantage for potential use within the higher education sector. In particular, we have illustrated its possible UK-wide applicability, as well as the ways in which it complements and offers additional benefits to existing area-level variables in this field.

It is, however, important to recognise that – as with all (individual and area-level) measures of disadvantage – our variable is not without limitations. The main criticism of area-level measures, such as POLAR and IMD, is the level of heterogeneity in family circumstances across neighbourhoods, meaning localities classified as relatively advantaged will still consist of pockets of deprivation and vice-versa. We have tried to mitigate this in the generation of our measure by relying upon the smallest geography level available. However, as our summary statistics in the appendix illustrate, there are still individuals living in output areas that we classify as disadvantaged who report that they have a parent with a higher education qualification or that is working in a professional occupation. Furthermore, while the Census is obligatory for households to complete, the data collected is still self-reported and could therefore be subject to the types of errors that exist for the parental education/occupation fields that are available in the HESA Student record via UCAS. Indeed, the ONS completed a Census Quality Survey between May and August 2011, in which volunteers were requested to participate in a face-to-face interview. Those who agreed were subsequently asked the same questions that they responded to approximately two to five months earlier, with the key difference being the mode of survey completion. Agreement rates were generally found to be lower (around two-thirds) for questions around occupation and qualification for reasons such as respondents giving different job titles or individuals struggling to remember the educational certificates they attained.[1]

Next steps in this programme of work are as follows. Firstly, we invite feedback and comments from all parts of the UK on the perceived usefulness of this measure. Should a positive response be received from our data users, we would then look to engage with user groups regarding incorporating this data into our collection and then distributing relevant extracts to practitioners and policymakers. We are aware that in this paper we have focused more on potential use of this variable among providers and policymakers. However, we appreciate there is also a need for suitable measures of socioeconomic disadvantage among the research community and in the analysis of policy initiatives/questions, as Jerrim (2020) highlights.[2] He cites the possible creation of an index that brings together both individual and area-level data to generate a continuous variable for socioeconomic disadvantage, which could offer the dual benefit of greater granularity and being less disclosive in nature. Consequently, if there is general approval for this UK-wide area-based measure, further exploration could involve examining whether an appropriate (UK and/or country-specific) index can be formulated that brings together our area-based measure with appropriate individual-level information, which we can then supply to data users.

Furthermore, with the 2021 Census having recently been submitted by citizens in England, Wales and Northern Ireland (with the Scottish Census taking place next year), we will have access to the latest area data over the course of the next few years. Throughout this paper, we have tried to alleviate any concerns over the applicability of our measure in present circumstances given it is based on 2011 Census data and change may occur over time. We have done this by illustrating that it captures areas that continued to report high levels of deprivation across the last decade. 2021 Census data will offer us the opportunity to update our measure and undertake a detailed investigation into the stability of the findings across the decade, particularly given that one of the aims of the Census is to maintain the ability to compare over time (e.g. through the use of consistent questions[3]).

To allow us to assess how the composition of students varies across POLAR, IMD and our measure, we have had to restrict our focus here to young entrants to higher education. However, as our measure is based on all adults aged over 16, it is potentially suitable in assisting mature entry into higher education as well (a current limitation of some area-based measures, such as POLAR). Should we receive a positive response on this measure we have developed, we would also be happy to introduce an analysis of mature students when updating our work using the 2021 Census.

Finally, we recognise that we are currently unable to explore higher education provision in further education colleges through our data, which is a particular issue when examining Scotland due to the sizeable proportion of higher education delivered through such establishments. This will also be an area that we shall aim to address in forthcoming years.[4]

Back: 5. Results Next: Appendix 1: English domiciled entrants

[4] While HESA do collect data on alternative providers, this process began in the middle part of the previous decade, preventing them from being included in this analysis. Future work will be able to incorporate this group into the exploration.