Official Statistics

Response to the consultation on the Below Average Resources measure

Published 23 January 2025

Introduction

The Department for Work and Pensions (DWP) is developing a new poverty measure named ‘Below Average Resources’ (BAR) based on the approach proposed by the Social Metrics Commission (SMC).

Below Average Resources: Developing a new poverty measure was first published on 18 January 2024.

As part of developing the new Official Statistics in Development a user consultation ran for 12 weeks, from 18 January to 11 April 2024. A response was planned for publication in July 2024, but this was postponed in accordance with General Election guidance. This response is now being published, alongside the Below Average Resources: FYE 2023 update.

Overview of responses

We received a total of 35 responses. These consisted of:

• 2 responses from government departments
• 16 responses from charities
• 4 responses from think tanks or other non-governmental research organisations
• 13 responses from various other sources (for example, academics, local authorities, members of the public, etc.)

Responses were received from both individuals and organisations. There were 31 responses on behalf of an organisation and 4 responses from individuals. All questions were optional, with varying levels of response across the consultation questions.

Summary of main responses

Overarching Social Metrics Commission approach

The response to the first question, on the value of the Overarching Measurement Framework proposed by the SMC, was overwhelmingly positive. Respondents were supportive of the measure in principle, though most responses were supportive subject to the further development work required. The vast majority of responses were supportive of the measure alongside, rather than in place of existing poverty measures, due to the different understanding each can provide and not wanting to lose historical time series and international comparability offered by existing measures.

All but three responses suggested the measure would add value for users. Respondents felt that with the additional components the measure offers a more comprehensive and detailed picture of poverty compared to existing measures, which would enable better targeting of support. However, several responses did highlight the increased complexity of the measure and challenge of interpreting the drivers behind changing rates across the components. Of the responses that were less convinced of the added value of the measure, one felt the measure was too complex, whilst another expressed preference for improving existing measures.

Respondents were asked what their development priorities would be. Frequently cited areas for development included: considering inclusion of a wider range of potential costs (including transport, social care) and providing more granular geographic and demographic breakdowns of data.

It was suggested that the successful development of the measure should be based on three broad objectives:

• methodological robustness with a settled and agreed methodology that stands up to the Code of Practice for Statistics
• adding new understanding of poverty and its causes over existing measures
• being accepted and used by government and other stakeholders, with equal weight alongside existing poverty measures

In the longer term, the majority of respondents believed there is value in DWP developing the Poverty Persistence and Lived Experience Indicators elements of the SMC framework, as well as consulting with groups with experience of poverty. Though respondents felt that further work and careful consideration would be required to develop the areas of focus for indicators and outlined the need to consider multiple disadvantage.

Detailed review of Total Resources Available (TRA) components

Assets

There was a mixed response to the question on whether liquid assets should be counted as an available resource within the TRA calculation. Of those that supported the principle, many identified potential risks and suggested areas for improvement. More work is needed on the definition of liquid assets, and whilst some respondents felt all types of assets should be treated equally to avoid further complexity, others suggested exempting different types of assets. Respondents were also mixed on the question of whether there should be a minimum savings buffer. There was broad consensus that the current basis used to convert liquid assets from a stock to a weekly flow measure was limited. Some respondents stated this was not realistic and one of the biggest concerns expressed with the current methodology.  

Debt

All responses were supportive of obligated debt repayments being counted as an inescapable cost within the TRA calculations. However, some agreement was in principle, dependent on quality of data available and if liquid assets were included.

In terms of defining obligated debt repayments, several responses suggested a broad conception including most types of debt. Most respondents did not agree that the amount borrowed for debt should be included as a source of income, in addition to being a cost. Reasons cited against were the principle of the measure based on ‘poverty now’ and that many debts may not be a consequence of capital borrowing.

Childcare costs

The consensus was that childcare costs should be included within the TRA calculation, but with several areas for further development highlighted, including considering coverage of available data, choices made around childcare and how government subsidies interact. On defining childcare costs, the variation in responses reflected the complexity of doing so. Several responses mentioned including actual spend on paid for childcare, whilst additional suggestions mentioned were including costs of provision of care element (for example, nappies and food).

When asked about the position that unmet need and informal care use are not possible to reflect, there was consensus that it was very difficult to reflect in the measure. Some potential options to explore were proposed, such as creating a model to reflect what each household would need, surveying or using Lived Experience Indicators instead.

When asked if there should be an adjustment for the quality of childcare (for example, including a cap on maximum cost incurred), the consensus was that there should not be, but that again more work is needed. Reasons for this included, costs being driven by location and availability of childcare, affected by age of child and it would be laden with assumptions.

Other suggestions proposed a before and after childcare costs version, and a potential need to adjust for age and disability.

Extra costs of disability

The overwhelming consensus was that the extra costs of disability should be counted as an inescapable cost within the Total Resources Available calculation with a number of suggestions for how the current approach should be improved. The disability benefits used as a proxy were considered not to be fully representative, with alternative suggestions including collecting evidence from disabled people directly or improving the data available in other ways and needing to consider that disabled people are not one homogenous group. Responses highlighted the complexity of attempting to define and measure the extra costs of disability with significant further work required in this area.

Housing costs

Responses to this section were fewer in number than others. However, there was broad consensus among respondents that mortgage capital repayments should be included in the Total Resources Available calculation.

Respondents were asked if they were to be included, whether overpayments of mortgage capital should be included. The general consensus was that they should not be, one argument being that it would add to data capture needs and add complexity, with minimal impact on the understanding of poverty.

When asked if the amount borrowed through a mortgage, the value of the asset gained by paying back a mortgage should be accounted for in the Total Resources Available calculation as an income, the consensus was that they should not be, and several respondents gave the reason that this is not a liquid asset.

When asked if there are any other housing associated inescapable costs that are not currently considered, insurance (structural, building and contents) was mentioned, as well as cost of repairs, costs of adaptions to make housing accessible and service charges. 

TRA components continued

When asked if there were any additional inescapable costs, not currently included, that should be included in the Total Resource Available calculation, the most common response was social care costs. This was followed by energy, transport, healthcare, caring, food and schooling. This points to a need to consider a guiding framework for which costs would be determined as inescapable and family-specific, versus those that would be more universal and captured through the equivalisation process as the measure is further developed.

Detailed review of poverty estimation method

Sharing Unit

When asked if respondents thought the revised approach to defining a sharing unit is required, there was broad agreement that it is. Several responses felt that it was an improvement over the existing household measure whilst still agreeing that it has its own limitations. A number of responses however, raised issues with the assumptions, such as the fact that the current definition assumes that those who are related share resources in a way that is equivalent for all its members, which is not always the case. Some respondents were also unsure on the value that it adds as poverty rates do not differ significantly to when they are measured at a household level.

Equivalisation

Respondents were asked if they agreed that a revised equivalisation approach is required and whilst there was broad consensus that the current approach is not an accurate reflection of the needs of different family compositions, many noted there is no easy solution.

Smoothing

When asked if the poverty threshold (if relative to social norms), should be smoothed so that changes to social norms are only reflected when sustained, mixed views were presented. Of those that agreed, most felt three years smoothing was appropriate. Some however, argued that one-year thresholds should be used instead due to ease of communication with the public and to make it easier to interpret year-on-year changes.

Determining the poverty threshold

Several respondents disagreed with the current approach to poverty threshold estimation (i.e. baselining to HBAI values in 2016 to 2017), identifying that it is currently trying to match previous estimates rather than being based on an assessment of needs. Many felt that maintaining the current 60% threshold from HBAI would be better as it has an evidence base and allows for better comparison over time. Others felt that the determination of a threshold will always be arbitrary no matter where it is set.

Overcrowding adjustment

Respondents were asked if any overcrowding adjustment should be applied after the determination of the smoothed median equivalised TRA and low resource threshold and the responses were mixed. Some respondents felt there should be an adjustment, albeit some do not necessarily agree with its current application. Several more thought there should not be an adjustment altogether, expressing that TRA should focus on measuring actual disposable cash and keep overcrowding separate.

Wider poverty measurement framework

Respondents were asked if poverty depth should be considered, to establish who is below the threshold and what value this adds for users. The overwhelming response was that it should be, and this would add value to the measure.

Respondents were asked if depth bands were an appropriate way to measure depth of poverty and if there were alternative approaches that should be considered. The consensus was that this would be appropriate. Alternative approaches suggested included banding using thresholds of median, as would be the median resources of someone in poverty as a ratio of the poverty line, looking at the poverty gap and having an equivalent line just above the poverty threshold, as an indicator of risk of falling into poverty.

Wider considerations

Respondents were asked for any views on using modelling and/or imputation for some elements as an alternative or complement to self-reported data. There was caution expressed in the responses and that it should only be done if the methodology is robust.

When asked if it was problematic that it may not be possible to produce a consistent historical time series for the BAR measure. There was broad consensus that breaking the time series to fully develop the BAR measure was unfortunate, but inevitable, and that developing the measure was more important than trying to maintain a consistent time series. Several respondents noted that current HBAI measures would still be able to provide an over-time view. Several respondents mentioned the importance of considering the wider context of FRS transformation, and how that may lead to revisions for all FRS based products, including BAR, in any case.

Conclusions and next steps

The user consultation highlighted overwhelming support for the value added by the Below Average Resources measure alongside existing poverty measures, once it is fully developed. However, the consultation responses also revealed the wide-ranging nature of uncertainty and differing views on how to develop the multiple components of the framework in practice across the detailed questions posed. Significant further development work is required to test approaches and develop the statistic into a robust and internally consistent measure.

Alongside this consultation response, we have published the Below Average Resources: FYE 2023 update with an updated baseline incorporating 2022 to 23 data. This provides a basis for us to conduct detailed further development work on each of these questions, incorporating the feedback received across the consultation and comparing proposed methodological changes against this baseline.

As the work is progressed, future developments will be communicated through the DWP statistical work programme and users will continue to be consulted on proposed developments in due course.