SEISA by HESA – a new area-based measure of deprivation

Existing area-based measures of deprivation have never really worked for higher education. HESA's Tej Nathwani introduces the new socio-economic index for small areas

Tej Nathwani is Principal Researcher at HESA

As a producer of official statistics, innovation is expected to be at the heart of what we do at HESA.

Over the last few years, we’ve therefore been working on creating a new UK-wide area based measure of deprivation, with today marking the release of the interactive maps/postcode search tool associated with our new variable – the Socioeconomic Index for Small Areas (SEISA).

In this blog, we take readers on a journey of what prompted us to conduct this work (the ‘why’) and our approach (the ‘how’). We conclude by outlining the ways in which you can get involved in its future development.

Innovation should be at the forefront of a statistical system that meets its users’ needs. (Mark Pont, Office for Statistics Regulation, 2017)

We’ve also included a new principle of innovation – it is all about the desire for improvement, the restless asking of questions like ‘why’ and ‘how’ (Ed Humpherson, Office for Statistics Regulation, 2017).

The remit of HESA

Back in the late 1980s and early 1990s, the growing levels of participation in higher education led to calls from policymakers for a more coherent approach to the collection and dissemination of statistics. Consequently, for the last three decades, HESA has been the body responsible for carrying out this function. Indeed, despite the fact that there has been great change within society over this time (including education becoming a devolved matter across the four nations), our core purpose remains to produce regular UK-wide official statistics on various aspects of the sector.

Publishing such outputs is accompanied with a duty to adhere to the Code of Practice for Statistics, which (as the above quotes indicate) actively encourages organisations to innovate and improve by, for example, plugging identified information gaps.

Find the gap

If one were to explore a recent report by the Social Mobility Commission (who are tasked with evaluating progress in social mobility across the UK) or from the Office for Statistics Regulation (OSR) on education statistics, they would find that the lack of a UK-wide deprivation field is explicitly discussed as a current limitation within available data. The main critique here is that existing variables – for varied reasons – don’t enable comparisons to be made across geographic domains. Comparability is often regarded as a key feature of quality statistics and can bring benefits such as allowing other nations to learn from the experiences of a country that has made progress towards achieving a particular policy objective. This potential advantage is also noted at the end of page 11 in a study published by the Education Policy Institute and the University of Oxford.

Sometimes, researchers will wish to perform a UK-wide analysis without any country/regional level comparisons. Most recently, this has been illustrated by the work of learned societies (e.g. in economics and chemistry) who want to better understand access to their subject area. Current UK-wide individual-level indicators on socioeconomic background that could be/have been used from HESA records in such instances, such as parental education or occupation, tend to have a larger proportion of missing data than area-based measures, which rely on the postcode of the student.

In addition, existing area-based measures commonly deployed in exploring disparities by socioeconomic background also have their drawbacks. Alongside the Indices of Deprivation being nation-specific, a long-standing criticism of the final index in each nation has been its inability to effectively capture deprivation in rural spots (as we discuss in our technical report). Also, the size of the areas used to form this measure can lead to it being difficult to pick up pockets of deprivation in otherwise less deprived localities. The impact of this is that students who could have benefited from activities designed to promote equal opportunity for all are overlooked.

Collectively, this evidence points to a need for a UK-wide deprivation measure that concurrently addresses some of the drawbacks of existing area-based indicators. The nature of our work means that we were well placed to tackle this challenge.

Our approach

Research can play a key role in the evolution of official statistics – something we note in our Research Strategy and which the OSR also point out in guidance on the development of new statistics. Based on the above, our aim was to create a measure that was:

  • UK-wide
  • Used smaller-sized areas than those used to create the Indices of Deprivation
  • Had near full coverage of the population/minimal missing data
  • Better captured rural areas

It was for this reason that when we began this work in early 2021, we relied on the latest Census (which was 2011 at the time). A key advantage of this collection is that it is mandatory for all households to complete once every ten years, with many of the resulting outputs being broadly or highly comparable across the UK. Furthermore, given data is available for areas which are smaller in population than those used for the Indices of Deprivation, this seemed the most suitable source to achieve our outlined objectives. With previous literature identifying education and occupation as key drivers of low income/deprivation, area-level data on these two factors were used to form our final measure.

Engaging with users

After releasing our first working paper in October 2021, we requested feedback from users (e.g. academics, widening participation practitioners, funding bodies etc) on our research. Positive comments were received, with the main points raised being that additional analysis exploring whether SEISA does proxy for deprivation as expected would be beneficial, as well as further comparisons with existing area-based measures. This resulted in us publishing an updated study in November 2022 covering these topics in greater detail (including how SEISA was generally picking up a higher proportion of rural areas in the bottom quintile).

It was only after producing these initial pieces of research that we circulated our first set of experimental statistics on the measure last April.

Simultaneously, to ensure the data can be accessed by the widest possible audience, we have been working on generating a set of interactive maps as a means of disseminating data on SEISA, with the Code of Practice stating that;

Producers should commit to improve data presentation, enhance insight, and better meet the needs of different types of users and potential users in the dissemination of their statistics and data.

Widening participation practitioners and academics from all regions of the UK were asked to take part in testing our webpages. Indeed, our final designs owe much to the very helpful comments we received from participants. For instance, as a result of the feedback received, we have constructed a summary page that answers key questions, such as how our measure was formed (with a worked example), how/where it can be applied and comparisons with existing measures used in the sector. Additionally, we have provided the dataset we used to produce our research, so that our findings can be replicated and to enable analysts to work with SEISA, alongside other area-level indicators.

Looking to the future

Area-based measures have often been criticised on the basis that they may not reflect the circumstances facing a given individual. That is, a person living in a deprived neighbourhood may well be from a high income household with a large level of resource. As we noted in a recent research piece though, empirical data suggests we cannot rule out that living in a deprived area inhibits an individual reaching their full potential irrespective of their family background. Should such findings continue to emerge, it is likely that area-level indicators will remain important in ensuring there is equal opportunity for all.

As noted in our ad-hoc bulletin last year, the intention is for SEISA to be an experimental statistic (or an “official statistic in development” as it is now referred to by the OSR) for 1-2 years. During this time, to help us understand the value of SEISA, we welcome user feedback on the measure. In particular, we’d be grateful if you could let us know;

  • The ways in which you are using (or may use) SEISA to support your activities/decision-making processes.
  • How useful SEISA has been (or could be) in your work and the reasons for this.
  • Whether you would endorse the measure being updated by HESA with data from the latest Census (2022 data from Scotland is due to be released later this year after which it would be possible to do this).

View the interactive maps/postcode search tool for SEISA on the HESA website from 9.30am.

Submit your thoughts on SEISA and the interactive maps at

Read HESA’s latest research releases, and if you would like to be kept updated on future publications, please sign-up to our mailing list.

2 responses to “SEISA by HESA – a new area-based measure of deprivation

  1. Thanx very much for this.

    Am I right in thinking that each SEISA area has about 200 residents (aged 16 to 74)?

  2. Having a little dig into the data, at a high-level it seems to have similar drawbacks to existing measures of deprivation that are used in a HE context.

    For example, only 0.5% of areas in London are in the most deprived 10% of areas in England on this measure and only 4.7% are in the most deprived 20%. This does not reflect the economic deprivation of London.

Leave a Reply